ODONTOGENESIS

Methods in
Molecular Biology 1922
Petros Papagerakis Editor
Odonto-
genesis
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
For further volumes:

http://www.springer.com/series/7651
Odontogenesis
Methods and Protocols
Edited by
Petros Papagerakis
School of Dentistry, University of Michigan, Ann Arbor, MI, USA
College of Dentistry, University of Saskatchewan, Saskatoon, SK, Canada
Editor
Petros Papagerakis
School of Dentistry
University of Michigan
Ann Arbor, MI, USA
College of Dentistry
University of Saskatchewan
Saskatoon, SK, Canada
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-9011-5 ISBN 978-1-4939-9012-2 (eBook)
https://doi.org/10.1007/978-1-4939-9012-2
Library of Congress Control Number: 2019930434
© Springer Science+Business Media, LLC, part of Springer Nature 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations
and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Humana Press imprint is published by the registered company Springer Science+Business Media, LLC, part of
Springer Nature.
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface
Odontogenesis is the complex process by which embryonic cells differentiate into oral
epithelia-derived ameloblasts that secrete enamel and cranial neural crest‐derived
mesenchyme, which form odontoblasts that produce dentin and cementoblasts that make
cementum. Different approaches are applied to study genetic and environmental regulatory
controls on odontogenesis that can have dramatic influences on dental phenotypes and
genotypes observed during normal development and diseases. The different methods used
have advantages and limitations, and this book aims to serve as a guide to future generation
of researchers working in the exciting field of odontogenesis. This book is divided into
6 parts and contains a total of 41 chapters.
Part I is focused on the establishment of dental cell lines and animal models that serve as
tools to understand the genetic and environmental controls of tooth development. Using
these models, researchers can further enhance our understanding on how signaling mole-
cules control all steps of tooth formation by coordinating cell proliferation, differentiation,
apoptosis, extracellular matrix synthesis, and mineral deposition. Chapters in this part
include protocols for isolation and characterization of both epithelial and mesenchymal
dental cells, establishment of stable cell lines, as well as in vivo cell lineage tracing and
classical tissue recombination assays using the kidney capsule model.
Part II is focused on dental stem cells and dental tissue regeneration. Tooth loss, caused
by dental diseases, trauma, or aging, is usually replaced by artificial materials which lack
many of the important biological characteristics of the natural tooth. Understanding the
mechanisms of stem cell differentiation toward dental phenotypes may provide the necessary
foundation that will lead to novel approaches for dental tissue regeneration and stem cell
therapies in the future. Methods described in this part will help researchers to further
elucidate the complex interactions and necessary conditions driving dental cell differentia-
tion. Chapters in this part detail methods on isolation, phenotypic characterization, expan-
sion, and differentiation protocols for dental stem cells as well as novel approaches for tissue
regeneration such as the use of multiwalled carbon nanotubes, peptides, or GelMA hydro-
gels. It also contains a protocol to study reparative dentinogenesis in vivo.
Part III is centered around methods to characterize gene and protein expression in
dental cells and tissues. New genes and their functions are continuously being discovered in
experimental studies using cell lines and animal models. This part provides the necessary
knowledge for successfully mapping RNA and protein expression in dental tissues. Detailed
protocols on immunofluorescence, in situ hybridization, immunohistochemistry (including
co-localization), and the use of LNA probes for detection of low amounts of RNA are
provided. Protocols for silver-albumin tissue staining and isolation of sibling proteins from
bone and dentin complete this part.
Part IV contains biochemistry and imaging protocols that are essential for characterizing
dental hard tissues. These methods, ranging from electron microscopy to micro-CT aided by
artificial intelligence, are critical for understanding gene function in transgenic and knock-
out mice models that may result in arrested tooth development and/or abnormal extracel-
lular matrix formation, maturation, and mineralization. Furthermore, protocols for
extraction and biochemical characterization of matrix proteins from enamel and dentin as
well as protocols for expression and purification of recombinant proteins are also provided.
v
vi Preface
Part V describes dental disease models focusing mainly on protocols to study dental
caries. Dental caries still cause a huge public health burden, and the in vitro and in vivo
models included here may help in developing new approaches for prevention, diagnosis, and
treatment of dental caries. Additional chapters include protocols of a rodent dental fluorosis
model and a method for 3D assessment of crown size and eruption space for third molars
allowing to study the effects of fluoride and third molar impaction, both commonly seen in
humans.
Part VI overviews protocols on genetics, epigenetics, and clinical studies to provide
foundation for clinical research in dentistry. Whole-genome linkage analysis, association
analysis of putative candidate genes, and whole-genome association approaches now offer
exciting opportunities to discover new key genes involved in human dental development.
This part contains protocols for next-generation sequencing, genetic and epigenetic studies,
and genome-wide association studies as well as clinical protocols for measurement of early
childhood caries and saliva and supragingival fluids and biofilm collection and subsequent
analyses.
Written in the highly successful Methods in Molecular Biology™ series format, chapters
include introductions to their respective topics; lists of the necessary materials and reagents;
step-by-step, readily reproducible laboratory protocols; and tips for troubleshooting and
avoiding expected pitfalls. Practical and easy-to-use Odontogenesis: Methods and Protocols
aims to guide researchers toward elucidating the secrets and mysteries of a fascinating and
unique organ, the tooth!
I am very grateful to all participants and their contributions to this volume. We all hope
this book will serve future generations of researchers in the field of odontogenesis in their
pathways to exciting discoveries.
Ann Arbor, MI, USA Petros Papagerakis

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
PART I PROTOCOLS FOR ESTABLISHMENT OF DENTAL CELLS LINES

ANDANIMAL MODELS
1 Microdissection and Isolation of Mouse Dental Epithelial Cells
of Continuously Growing Mouse Incisors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Hidemitsu Harada and Keishi Otsu
2 Establishment of an Immortalized Mouse Bmp2 Knockout Dental
Papilla Mesenchymal Cell Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Wen’an Xu and Shuo Chen
3 Establishment of Stable Cell Lines from Primary Human Dental Pulp
Stem Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Elizabeth Guirado, Youbin Zhang, and Anne George
4 Isolation of Dental Stem Cell-Enriched Populations from Continuously
Growing Mouse Incisors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Anamaria Balic
5 Application of Cell Lineage Tracing Combined with Immunofluorescence
in the Study of Dentinogenesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Yan Jing, Chaoyuan Li, and Jian Q. Feng
6 Tissue Recombination and Kidney Capsule Transplantation Assays
for the Study of Epithelial-Mesenchymal Interactions . . . . . . . . . . . . . . . . . . . . . . . 49
Lucia Jimenez-Rojo and Thimios A. Mitsiadis
PART II PROTOCOLS FOR DENTAL STEM CELLS AND TISSUE REGENERATION
7 Dental Mesenchymal Stem Cells: Dental Pulp Stem Cells, Periodontal

Ligament Stem Cells, Apical Papilla Stem Cells, and Primary Teeth
Stem Cells—Isolation, Characterization, and Expansion for Tissue
Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Mey Al-Habib and George T. -J. Huang
8 Phenotypic Identification of Dental Pulp Mesenchymal Stem/Stromal
Cells Subpopulations with Multiparametric Flow Cytometry . . . . . . . . . . . . . . . . . 77
Maxime Ducret, Jean-Christophe Farges, Marielle Pasdeloup,
Emeline Perrier-Groult, Andreas Mueller, Frédéric Mallein-Gerin,
and Hugo Fabre
9 Dental Pulp Stem Cells: Isolation, Characterization, Expansion,
and Odontoblast Differentiation for Tissue Engineering . . . . . . . . . . . . . . . . . . . . . 91
Qing Dong, Yuanyuan Wang, Fatemeh Mohabatpour, Li Zheng,
Silvana Papagerakis, Daniel Chen, and Petros Papagerakis
vii
viii Contents
10 In Vitro Analysis of Intramolecular Signaling Events in PDLSCs

Using Confocal and TIRF Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Annette Merkel and Anne George
11 A Mouse Model to Study Reparative Dentinogenesis. . . . . . . . . . . . . . . . . . . . . . . . 111
R. C. Babb, D. Chandrasekaran, L. K. Zaugg, and P. T. Sharpe
12 Multiwalled Carbon Nanotubes for Dental Applications . . . . . . . . . . . . . . . . . . . . . 121
Petros Kechagioglou, Eleftherios Andriotis, Petros Papagerakis,
and Silvana Papagerakis
13 Peptide-Mediated Biomimetic Regrowth of Human Enamel In Situ . . . . . . . . . . 129
Kaushik Mukherjee, Qichao Ruan, and Janet Moradian-Oldak
14 Bioengineering Tooth Bud Constructs Using GelMA Hydrogel . . . . . . . . . . . . . . 139
Elizabeth E. Smith and Pamela C. Yelick
15 Whole-Mount In Situ Hybridization of Mouse Embryos
Using DIG-Labeled RNA Probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Jingyi Wu and Xiaofang Wang
PART III PROTOCOLS FOR STUDYING GENE AND PROTEIN EXPRESSION
16 In Situ Hybridization on Mouse Paraffin Sections Using DIG-Labeled

RNA Probes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
Jingyi Wu, Jian Q. Feng, and Xiaofang Wang
17 Methods for In Situ Protein Visualization in Dental Mineralized Tissues . . . . . . 173
D. Hotton, A. Berdal, and A. Bolaños
18 In Situ Hybridization in Mineralized Tissues: The Added Value
of LNA Probes for RNA Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
G. Lignon, D. Hotton, A. Berdal, and A. Bolaños
19 Immunofluorescence Procedure for Developing Enamel Tissues . . . . . . . . . . . . . . 191
Xu Yang and Elia Beniash
20 Silver-Albumin Tissue Staining Protocol to Visualize Odontogenesis
in Whole Embryos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Julia C. Boughner and David M. L. Cooper
21 Isolation of SIBLING Proteins from Bone and Dentin Matrices . . . . . . . . . . . . . . 211
22 Immunohistochemical Co-Localization of Amelogenin and Ameloblastin
in Developing Enamel Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
Rucha Arun Bapat and Janet Moradian-Oldak
23 The Expression and Purification of Recombinant Mouse Ameloblastin
in E. coli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Jingtan Su, Rucha Arun Bapat, and Janet Moradian-Oldak
PART IV PROTOCOLS FOR BIOCHEMISTRY AND IMAGING

24 Protocols for Studying Formation and Mineralization
of Dental Tissues In Vivo: Extraction Protocol for Isolating Dentin
Matrix Proteins from Developing Teeth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
Yasuo Yamakoshi, Jan C.-C. Hu, Mari M. Saito, and James P. Simmer
Contents ix
25 Purification of Developing Enamel Matrix Proteins Using Preparative

SDS-PAGE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Steven J. Brookes and Claire M. Gabe
26 Using ImageJ (FIJI) to Analyze and Present X-Ray CT Images of Enamel . . . . . 267
Steven J. Brookes
27 Scanning Electron Microscopy (SEM) Methods for Dental Enamel . . . . . . . . . . . 293
Steinar Risnes, Muhammad Saeed, and Amer Sehic
28 Microcomputed Tomography Imaging in Odontogenesis Studies. . . . . . . . . . . . . 309
Kostas Verdelis and Phil Salmon
29 Transmission Electron Microscopy (TEM) and Scanning Electron
Microscopy (SEM) for the Examination of Dental Hard Tissues . . . . . . . . . . . . . . 325
Victor E. Arana-Chavez and Leticia S. Castro-Filice
PART V PROTOCOLS TO STUDY DENTAL DISEASES
30 Rodent Dental Fluorosis Model: Extraction of Enamel Organ

from Rat Incisors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
M. Suzuki and J. D. Bartlett
31 Three-Dimensional Assessment of Crown Size and Eruption Space
for Developing Third Molars: Data Collection Techniques Based
on Cone-Beam Computed Tomography (CBCT) . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
D. F. Marchiori, G. V. Packota, and J. C. Boughner
32 Protocols to Study Dental Caries In Vitro: Microbial Caries Models . . . . . . . . . . 357
Bennett T. Amaechi, Livia M. A. Tenuta, Antonio P. Ricomini Filho,
and Jaime A. Cury
33 In Vitro Caries Models for the Assessment of Novel Restorative Materials . . . . . 369
Basma Sulaiman Ghandourah, Anna Lefkelidou, Raed Said,
Xanthippi Chatzistavrou, Susan Flannagan, Carlos Gonzáles-Cabezas,
Christopher J. Fenno, Li Zheng, Silvana Papagerakis,
and Petros Papagerakis
34 Protocols to Study Dental Caries In Vitro: pH Cycling Models. . . . . . . . . . . . . . . 379
Bennett T. Amaechi
35 In Vivo Rodent Models for Studying Dental Caries and Pulp Disease . . . . . . . . . 393
June Hsiao, Yuanyuan Wang, Li Zheng, Ruirui Liu, Raed Said,
Lubomir Hadjiyski, Heekon Cha, Tatiana Botero, Xanthippi Chatzistavrou,
Qing Dong, Silvana Papagerakis, and Petros Papagerakis
PART VI PROTOCOLS FOR GENETIC, EPIGENETIC AND CLINICAL STUDIES
36 Protocol GenoDENT: Implementation of a New NGS Panel

for Molecular Diagnosis of Genetic Disorders with Orodental
Involvement. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
Tristan Rey, Julien Tarabeux, Bénédicte Gerard, Marion Delbarre,
Antony Le Béchec, Corinne Stoetzel, Megana Prasad,
Virginie Laugel-Haushalter, Marzena Kawczynski, Jean Muller,
Jamel Chelly, Hélène Dollfus, Marie-Cécile Manière,
and Agnès Bloch-Zupan
x Contents
37 Protocols for Genetic and Epigenetic Studies of Rare Diseases

Affecting Dental Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
Bruna Rabelo Amorim, Pollyanna Almeida Costa dos Santos,
Caroline Lourenço de Lima, Denise Carleto Andia,
Juliana Forte Mazzeu, and Ana Carolina Acevedo
38 Protocols, Methods, and Tools for Genome-Wide Association
Studies (GWAS) of Dental Traits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
Cary S. Agler, Dmitry Shungin, Andrea G. Ferreira Zandoná,
Paige Schmadeke, Patricia V. Basta, Jason Luo, John Cantrell,
Thomas D. Pahel Jr., Beau D. Meyer, John R. Shaffer, Arne S. Schaefer,
Kari E. North, and Kimon Divaris
39 Measurement of Early Childhood Oral Health for Research Purposes:
Dental Caries Experience and Developmental Defects of the Enamel
in the Primary Dentition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511
Jeannie Ginnis, Andrea G. Ferreira Zandoná, Gary D. Slade,
John Cantrell, Mikafui E. Antonio, Bhavna T. Pahel, Beau D. Meyer,
Poojan Shrestha, Miguel A. Simancas-Pallares, Ashwini R. Joshi,
and Kimon Divaris
40 The Supragingival Biofilm in Early Childhood Caries: Clinical
and Laboratory Protocols and Bioinformatics Pipelines Supporting
Metagenomics, Metatranscriptomics, and Metabolomics Studies
of the Oral Microbiome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
Kimon Divaris, Dmitry Shungin, Adaris Rodrı́guez-Cortés,
Patricia V. Basta, Jeff Roach, Hunyong Cho, Di Wu,
Andrea G. Ferreira Zandoná, Jeannie Ginnis, Sivapriya Ramamoorthy,
Jason M. Kinchen, Jakub Kwintkiewicz, Natasha Butz, Apoena A. Ribeiro,
and M. Andrea Azcarate-Peril
41 Saliva and Gingival Crevicular Fluid (GCF) Collection
for Biomarker Screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Petros Papagerakis, Li Zheng, Doohak Kim, Raed Said,
Amber A. Ehlert, Kevin K. M. Chung, and Silvana Papagerakis
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
Contributors
ANA CAROLINA ACEVEDO Oral Histopathology Laboratory, Department of Dentistry, Health

Sciences Faculty, University of Brası́lia (UnB), Brası́lia, Brazil
CARY S. AGLER Oral and Craniofacial Health Sciences, UNC School of Dentistry,
University of North Carolina-Chapel Hill, Chapel Hill, NC, USA
MEY AL-HABIB Faculty of Dentistry, Department of Endodontics, King Abdulaziz
University, Jeddah, Saudi Arabia
BENNETT T. AMAECHI Department of Comprehensive Dentistry, University of Texas Health
Science Center at San Antonio, San Antonio, TX, USA
BRUNA RABELO AMORIM Oral Histopathology Laboratory, Department of Dentistry, Health
Sciences Faculty, University of Brası́lia (UnB), Brası́lia, Brazil
DENISE CARLETO ANDIA Department of Epigenetics, Dental Research Division, Health
Science Institute, Paulista University (UNIP), São Paulo, SP, Brazil
ELEFTHERIOS ANDRIOTIS Laboratory of Organic Chemical Technology, Department of
Chemistry, Aristotle University of Thessaloniki, Thessaloniki, Greece
MIKAFUI E. ANTONIO Oral and Craniofacial Health Sciences, UNC School of Dentistry,
VICTOR E. ARANA-CHAVEZ Laboratory of Oral Biology, Department of Biomaterials and
Oral Biology, School of Dentistry, University of São Paulo, São Paulo, SP, Brazil
M. ANDREA AZCARATE-PERIL Center for Gastrointestinal Biology and Disease, Division
of Gastroenterology and Hepatology, and UNC Microbiome Core, Department of Medicine,
School of Medicine, University of North Carolina, Chapel Hill, NC, USA
R. C. BABB Department of Craniofacial Development and Stem Cell Biology, Centre for
Craniofacial and Regenerative Biology (CCRB), Dental Institute, King’s College London,
London, UK
ANAMARIA BALIC Research Program in Developmental Biology, Institute of Biotechnology,
University of Helsinki, Helsinki, Finland
RUCHA ARUN BAPAT Center for Craniofacial Molecular Biology, Herman Ostrow School of
Dentistry, University of Southern California, Los Angeles, CA, USA
JOHN D. BARTLETT Division of Biosciences, Ohio State University, College of Dentistry,
Columbus, OH, USA
PATRICIA V. BASTA Department of Epidemiology, Gillings School of Global Public Health,
University of North Carolina-Chapel Hill, Chapel Hill, NC, USA; Biospecimen Core
Processing Facility, Gillings School of Global Public Health, University of North Carolina-
Chapel Hill, Chapel Hill, NC, USA
ELIA BENIASH Department of Oral Biology, School of Dental Medicine, University of
Pittsburgh, Pittsburgh, PA, USA; Department of Bioengineering, Center for Craniofacial
Regeneration, McGowan Institute for Regenerative Medicine, Swanson School of
Engineering, University of Pittsburgh, Pittsburgh, PA, USA
ARIANE BERDAL Molecular Oral Pathophysiology, Cordeliers Research Center, UMRS 1138
INSERM, Paris-Descartes, Pierre-et-Marie-Curie, Paris-Diderot Universities, Paris,
France
xi
xii Contributors
AGNÈS BLOCH-ZUPAN Faculté de Chirurgie Dentaire, Université de Strasbourg, Strasbourg,

France; Hôpitaux Universitaires de Strasbourg, Pôle de Médecine et Chirurgie Bucco-
Dentaires, Centre de Référence des Maladies Rares Orales et Dentaires (CRMR, Reference
Center for Rare Oral Diseases), O-Rares, Strasbourg, France; Institut de Génétique et de
Biologie Moléculaire and Cellulaire, Centre Européen de Recherche en Biologie et en Mé
decine, CNRS UMR7104, INSERM U1258, Université de Strasbourg, Strasbourg,
Illkirch, France; Université de Strasbourg Institut d’Etudes Avancées USIAS, Strasbourg,
France; University College London, Eastman Dental Institute, London, UK
ALBA BOLAÑOS Molecular Oral Pathophysiology, Cordeliers Research Center, UMRS 1138
INSERM. Paris-Descartes, Pierre-et-Marie-Curie, Paris-Diderot Universities, Paris,
France
TATIANA BOTERO Department of Cariology, Restorative Sciences and Endodontics, School of
Dentistry, University of Michigan, Ann Arbor, MI, USA
JULIA C. BOUGHNER Department of Anatomy, Physiology and Pharmacology, College of
Medicine, University of Saskatchewan, Saskatoon, SK, Canada
STEVEN J. BROOKES Division of Oral Biology, School of Dentistry, University of Leeds, Leeds,
UK
NATASHA BUTZ Microbiome Core Facility, Department of Cell Biology and Physiology, School
of Medicine, University of North Carolina-Chapel Hill, Chapel Hill, NC, USA; Division of
Gastroenterology and Hepatology, School of Medicine, Department of Medicine, University
of North Carolina-Chapel Hill, Chapel Hill, NC, USA
JOHN CANTRELL Oral and Craniofacial Health Sciences, UNC School of Dentistry,
LETICIA S. CASTRO-FILICE Laboratory of Oral Biology, Department of Biomaterials and
Oral Biology, School of Dentistry, University of São Paulo, São Paulo, SP, Brazil
HEEKON CHA Department of Radiology, School of Medicine, University of Michigan, Ann
Arbor, MI, USA
D. CHANDRASEKARAN Department of Craniofacial Development and Stem Cell Biology,
Centre for Craniofacial and Regenerative Biology (CCRB), Dental Institute, King’s
College London, London, UK
XANTHIPPI CHATZISTAVROU Department of Orthodontics and Pediatric Dentistry, School of
JAMEL CHELLY Hôpitaux Universitaires de Strasbourg, Laboratoires de diagnostic génétique,
Institut de Génétique Médicale d’Alsace, Strasbourg, France; Institut de Génétique et de
Biologie Moléculaire and Cellulaire, Centre Européen de Recherche en Biologie et en Mé
decine, CNRS UMR7104, INSERM U964, Université de Strasbourg, Strasbourg, France
DANIEL CHEN Department of Mechanical Engineering, College of Engineering, University
of Saskatchewan, Saskatoon, SK, Canada
SHUO CHEN Department of Developmental Dentistry, School of Dentistry, The University of
Texas Health Science at San Antonio, San Antonio, TX, USA
HUNYONG CHO Department of Biostatistics, Gillings School of Global Public Health,
KEVIN K. M. CHUNG Department of Orthodontics and Pediatric Dentistry, School of
DAVID M. L. COOPER Department of Anatomy, Physiology and Pharmacology, College of
Medicine, University of Saskatchewan, Saskatoon, SK, Canada
JAIME A. CURY Department of Physiological Sciences, Piracicaba Dental School, University
of Campinas, Piracicaba, SP, Brazil
Contributors xiii
CAROLINE LOURENÇO DE LIMA Oral Histopathology Laboratory, Department of Dentistry,

Health Sciences Faculty, University of Brası́lia (UnB), Brası́lia, Brazil; Molecular
Pharmacology Laboratory, Department of Pharmaceutics Science, Health Sciences Faculty,
University of Brası́lia (UnB), Brası́lia, Brazil
MARION DELBARRE Hôpitaux Universitaires de Strasbourg, Laboratoires de diagnostic géné
tique, Institut de Génétique Médicale d’Alsace, Strasbourg, France
KIMON DIVARIS Department of Epidemiology, Gillings School of Global Public Health,
University of North Carolina-Chapel Hill, Chapel Hill, NC, USA; Department of
Pediatric Dentistry, UNC School of Dentistry, University of North Carolina-Chapel Hill,
Chapel Hill, NC, USA
HÉLÈNE DOLLFUS Laboratoire de Génétique Médicale, INSERM UMRS1112, Institut de
Génétique Médicale d’Alsace, FMTS, Université de Strasbourg, Strasbourg, France
QING DONG Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
University of Michigan, Ann Arbor, MI, USA; Department of Pediatric Dentistry, College
of Stomatology, North China University of Science and Technology, Tangshan, China
POLLYANNA ALMEIDA COSTA DOS SANTOS Oral Histopathology Laboratory, Department of
Dentistry, Health Sciences Faculty, University of Brası́lia (UnB), Brası́lia, Brazil;
Universidade Estadual de Ciências da Saúde de Alagoas, Maceio, Alagoas, Brazil
MAXIME DUCRET Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique, UMR5305
CNRS/Université Lyon 1, UMS3444 BioSciences Gerland-Lyon Sud, Lyon, France; Faculté
d’Odontologie, Université de Lyon, Université Lyon 1, Lyon, France; Hospices Civils de
Lyon, Service de Consultations et Traitements Dentaires, Lyon, France
AMBRE A. EHLERT Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
University of Michigan, Ann Arbor, MI, USA
HUGO FABRE Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique, UMR5305
CNRS/Université Lyon 1, UMS3444 BioSciences Gerland-Lyon Sud, Lyon, France;
Laboratory of Regenerative Technologies, Department of Biomedical Engineering,
University of Basel, Basel, Switzerland
JEAN-CHRISTOPHE FARGES Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique,
UMR5305 CNRS/Université Lyon 1, UMS3444 BioSciences Gerland-Lyon Sud, Lyon,
France; Faculté d’Odontologie, Université de Lyon, Université Lyon 1, Lyon, France;
Hospices Civils de Lyon, Service de Consultations et Traitements Dentaires, Lyon, France
JIAN Q. FENG Department of Biomedical Sciences and Center for Craniofacial Research
and Diagnosis, Texas A&M University College of Dentistry, Dallas, TX, USA
CHRISTOPHER J. FENNO Department of Biologic and Materials Sciences, School of Dentistry,
ANDREA G. FERREIRA ZANDONÁ Department of Comprehensive Dentistry, Tufts University
School of Dental Medicine, Tufts University, Boston, MA, USA
SUSAN FLANNAGAN Department of Cariology, Restorative Sciences and Endodontics, School
of Dentistry, University of Michigan, Ann Arbor, MI, USA
CLAIRE M. GABE Division of Oral Biology, School of Dentistry, University of Leeds, Leeds, UK
ANNE GEORGE Department of Oral Biology, College of Dentistry, University of Illinois at
Chicago, Chicago, IL, USA
BÉNÉDICTE GERARD Hôpitaux Universitaires de Strasbourg, Laboratoires de diagnostic géné
JEANNIE GINNIS Department of Pediatric Dentistry, UNC School of Dentistry, University of
North Carolina-Chapel Hill, Chapel Hill, NC, USA
xiv Contributors
CARLOS GONZÁLEZ-CABEZAS Department of Cariology, Restorative Sciences and

Endodontics, School of Dentistry, University of Michigan, Ann Arbor, MI, USA
ELIZABETH GUIRADO Department of Oral Biology, College of Dentistry, University of Illinois
at Chicago, Chicago, IL, USA
LUBOMIR HADJIYSKI Department of Radiology, School of Medicine, University of Michigan,
Ann Arbor, MI, USA; Comprehensive Cancer Center, University of Michigan, Ann Arbor,
MI, USA
HIDEMITSU HARADA Division of Developmental Biology and Regenerative Medicine,
Department of Anatomy, Iwate Medical University, Shiwa-gun, Iwate, Japan
DOMINIQUE HOTTON Centre de Recherche des Cordeliers, INSERM UMR_S1138, Equipe
Physiopathologie Orale Moléculaire, University Paris-Diderot, Paris, France; Molecular
Oral Pathophysiology, Cordeliers Research Center, UMRS 1138 INSERM, Paris-Descartes,
Pierre-et-Marie-Curie, Paris-Diderot Universities, Paris, France
JUNE HSIAO Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
GEORGE T.-J. HUANG Department of Bioscience Research, University of Tennessee Health
Science Center, College of Dentistry, Memphis, TN, USA
JAN C.-C. HU Department of Biologic and Materials Sciences, University of Michigan
School of Dentistry, Ann Arbor, MI, USA
LUCIA JIMENEZ-ROJO Faculty of Medicine, Orofacial Development and Regeneration,
Institute of Oral Biology, Center of Dental Medicine, ZZM, University of Zurich, Zurich,
Switzerland
YAN JING Department of Orthodontics, Texas A&M University College of Dentistry, Dallas,
TX, USA
ASHWINI R. JOSHI Oral and Craniofacial Health Sciences, UNC School of Dentistry,
MARZENA KAWCZYNSKI Faculté de Chirurgie Dentaire, Université de Strasbourg, Strasbourg,
France; Hôpitaux Universitaires de Strasbourg, Pôle de Médecine et Chirurgie Bucco-
Dentaires, Centre de Référence des Maladies Rares Orales et Dentaires (CRMR, Reference
Center for Rare Oral Diseases), O-Rares, Strasbourg, France
PETROS KECHAGIOGLOU Department of Surgery, College of Medicine, University of
Saskatchewan, Saskatoon, SK, Canada; College of Dentistry, University of Saskatchewan,
DOOHAK KIM Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
JASON M. KINCHEN Metabolon, Inc., Durham, NC, USA
JAKUB M. KWINTKIEWICZ Division of Gastroenterology and Hepatology, Department of
Medicine, Microbiome Core Facility, Center for Gastrointestinal Biology and Disease,
School of Medicine, University of North Carolina, Chapel Hill, NC, USA
VIRGINIE LAUGEL-HAUSHALTER Laboratoire de Génétique Médicale, INSERM UMRS1112,
Institut de Génétique Médicale d’Alsace, FMTS, Université de Strasbourg, Strasbourg,
France
ANTONY LE BÉCHEC Hôpitaux Universitaires de Strasbourg, IRC, Institut Régional du
Cancer, Strasbourg, France
ANNA LEFKELIDOU Department of Orthodontics and Pediatric Dentistry, School of
Dentistry, University of Michigan, Ann Arbor, MI, USA; Department of Pediatric
Dentistry, Aristotle University of Thessaloniki, Thessaloniki, Greece
Contributors xv
CHAOYUAN LI Department of Biomedical Sciences, Texas A&M University College of

Dentistry, Dallas, TX, USA; Shanghai Engineering Research Center of Tooth Restoration
and Regeneration, Department of Oral Implant, School of Stomatology, Tongji University,
Shanghai, People’s Republic of China
G. LIGNON Molecular Oral Pathophysiology, Cordeliers Research Center, UMRS 1138
INSERM, Paris Descartes, Pierre-et-Marie-Curie, Paris-Diderot Universities, Paris,
France
RUIRUI LIU Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
University of Michigan, Ann Arbor, MI, USA; Department of Prosthodontics, School &
Hospital of Stomatology, Xi’an Jiaotong University, Jiaotong, China
JASON LUO Lineberger Comprehensive Cancer Center, School of Medicine, University of
North Carolina, Chapel Hill, NC, USA; Mammalian Genotyping Core, University of
North Carolina, Chapel Hill, NC, USA
FRÉDÉRIC MALLEIN-GERIN Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique,
France
MARIE-CÉCILE MANIÈRE Faculté de Chirurgie Dentaire, Université de Strasbourg,
Strasbourg, France; Hôpitaux Universitaires de Strasbourg, Pôle de Médecine et Chirurgie
Bucco-Dentaires, Centre de Référence des Maladies Rares Orales et Dentaires (CRMR,
Reference Center for Rare Oral Diseases O-Rares), Strasbourg, France
DENVER F. MARCHIORI Department of Anatomy, Physiology, and Pharmacology, College
of Medicine, University of Saskatchewan, Saskatoon, SK, Canada
JULIANA FORTE MAZZEU Laboratory of Medical Genetics, Faculty of Medicine, University
of Brası́lia (UnB), Brası́lia, Brazil
ANNETTE MERKEL Department of Oral Biology, College of Dentistry, The University of
Illinois at Chicago, Chicago, IL, USA
BEAU D. MEYER Department of Pediatric Dentistry, UNC School of Dentistry, University
THIMIOS A. MITSIADIS Faculty of Medicine, Orofacial Development and Regeneration,
Institute of Oral Biology, Center of Dental Medicine, ZZM, University of Zurich, Zurich,
Switzerland
FATEMEH MOHABATPOUR College of Dentistry, University of Saskatchewan, Saskatoon, SK,
Canada; Department of Mechanical Engineering, College of Engineering, University
of Saskatchewan, Saskatoon, SK, Canada
JANET MORADIAN-OLDAK Center for Craniofacial Molecular Biology, Herman Ostrow
School of Dentistry, University of Southern California, Los Angeles, CA, USA; Department
of Biomedical Engineering, University of Southern California, Los Angeles, CA, USA
ANDREAS MUELLER Laboratory of Regenerative Technologies, Department of Biomedical
Engineering, University of Basel, Basel, Switzerland
KAUSHIK MUKHERJEE Center for Craniofacial Molecular Biology, Herman Ostrow School
of Dentistry, University of Southern California, Los Angeles, CA, USA
JEAN MULLER Hôpitaux Universitaires de Strasbourg, Laboratoires de diagnostic génétique,
Institut de Génétique Médicale d’Alsace, Strasbourg, France; Laboratoire de Génétique Mé
dicale, INSERM UMRS1112, Institut de Génétique Médicale d’Alsace, FMTS, Université
de Strasbourg, Strasbourg, France
KARI E. NORTH Department of Epidemiology, Gillings School of Global Public Health,
University of North Carolina, Chapel Hill, NC, USA; Carolina Center for Genome
Sciences, University of North Carolina-Chapel Hill, Chapel Hill, NC, USA
xvi Contributors
KEISHI OTSU Division of Developmental Biology and Regenerative Medicine, Department

of Anatomy, Iwate Medical University, Shiwa-gun, Iwate, Japan
GARNET V. PACKOTA College of Dentistry, University of Saskatchewan, Saskatoon, SK,
Canada
BHAVNA T. PAHEL Department of Pediatric Dentistry, UNC School of Dentistry, University
THOMAS D. PAHEL JR. Oral and Craniofacial Health Sciences, UNC School of Dentistry,
PETROS PAPAGERAKIS School of Dentistry, University of Michigan, Ann Arbor, MI, USA;
College of Dentistry, University of Saskatchewan, Saskatoon, SK, Canada
SILVANA PAPAGERAKIS Department of Surgery, College of Medicine, University of
Saskatchewan, Saskatoon, SK, Canada; Department of Otolaryngology Head and Neck
Surgery, School of Medicine, University of Michigan, Ann Arbor, MI, USA; Toxicology
Interdisciplinary Program, University of Saskatchewan, Saskatoon, SK, Canada;
Biomedical Engineering, University of Saskatchewan, Saskatoon, SK, Canada
MARIELLE PASDELOUP Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique,
France
EMELINE PERRIER-GROULT Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique,
France
MEGANA PRASAD Laboratoire de Génétique Médicale, INSERM UMRS1112, Institut de
SIVAPRIYA RAMAMOORTHY Metabolon, Inc., Durham, NC, USA
TRISTAN REY Faculté de Chirurgie Dentaire, Université de Strasbourg, Strasbourg, France;
Hôpitaux Universitaires de Strasbourg, Laboratoires de diagnostic génétique, Institut de Gé
nétique Médicale d’Alsace, Strasbourg, France
APOENA A. RIBEIRO Department of Diagnostic Sciences, UNC School of Dentistry,
ANTONIO P. RICOMINI FILHO Department of Physiological Sciences, Piracicaba Dental
School, University of Campinas, Piracicaba, SP, Brazil
STEINAR RISNES Faculty of Dentistry, Institute of Oral Biology, University of Oslo, Oslo,
Norway
JEFF ROACH Research Computing, University of North Carolina-Chapel Hill, Chapel Hill,
NC, USA
ADARIS RODRÍGUEZ-CORTÉS Department of Epidemiology, Gillings School of Global Public
Health, University of North Carolina-Chapel Hill, Chapel Hill, NC, USA; Biospecimen
Core Processing Facility, Gillings School of Global Public Health, University of North
Carolina-Chapel Hill, Chapel Hill, NC, USA
QICHAO RUAN Ormco Corporation, Glendora, USA; Center for Craniofacial Molecular
Biology, Herman Ostrow School of Dentistry, University of Southern California,
Los Angeles, CA, USA
MUHAMMAD SAEED Faculty of Dentistry, Institute of Oral Biology, University of Oslo, Oslo,
Norway
RAED SAID College of Dentistry, University of Saskatchewan, Saskatoon, SK, Canada;
Department of Anatomy and Cell Biology, College of Medicine, University of Saskatchewan,
Contributors xvii
MARI M. SAITO Department of Biochemistry and Molecular Biology, School of Dental

Medicine, Tsurumi University, Yokohama, Japan
PHIL SALMON Department of Restorative Dentistry/Comprehensive Care, School of Dental
Medicine, University of Pittsburgh, Pittsburgh, PA, USA; Bruker microCT, Kontich,
Belgium
ARNE S. SCHAEFER Department of Periodontology, Institute of Dental, Oral and Maxillary
Medicine, Charité—University Medicine Berlin, Berlin, Germany
PAIGE SCHMADEKE Department of Epidemiology, Gillings School of Global Public Health,
University of North Carolina-Chapel Hill, Chapel Hill, NC, USA; Biospecimen Core
Processing Facility, Gillings School of Global Public Health, University of North Carolina-
Chapel Hill, Chapel Hill, NC, USA
AMER SEHIC Faculty of Dentistry, Institute of Oral Biology, University of Oslo, Oslo, Norway
JOHN R. SHAFFER Center for Craniofacial and Dental Genetics, School of Dental Medicine,
University of Pittsburgh, Pittsburgh, PA, USA; Department of Oral Biology, School of
Dental Medicine, University of Pittsburgh, Pittsburgh, PA, USA; Department of Human
Genetics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA, USA
PAUL T. SHARPE Department of Craniofacial Development and Stem Cell Biology, Centre
for Craniofacial and Regenerative Biology (CCRB), Dental Institute, King’s College
London, London, UK
POOJAN SHRESTHA Department of Pediatric Dentistry, UNC School of Dentistry, University
DMITRY SHUNGIN Department of Odontology, Umeå University, Umeå, Sweden; Broad
Institute of the Massachusetts Institute of Technology and Harvard University, Cambridge,
MA, USA
MIGUEL A. SIMANCAS-PALLARES Oral and Craniofacial Health Sciences, UNC School of
Dentistry, University of North Carolina-Chapel Hill, Chapel Hill, NC, USA
JAMES P. SIMMER Department of Biologic and Materials Sciences, University of Michigan
School of Dentistry, Ann Arbor, MI, USA
GARY D. SLADE Department of Dental Ecology, UNC School of Dentistry, University of
North Carolina-Chapel Hill, Chapel Hill, NC, USA
ELIZABETH E. SMITH Department of Cell, Molecular, and Developmental Biology, Sackler
School of Graduate Biomedical Sciences, Tufts University School Medicine, Boston, MA,
USA; Department of Orthodontics, Tufts University School of Dental Medicine, Boston,
MA, USA
CORINNE STOETZEL Laboratoire de Génétique Médicale, INSERM UMRS1112, Institut de
JINGTAN SU Center for Craniofacial Molecular Biology, Herman Ostrow School of Dentistry,
University of Southern California, Los Angeles, CA, USA
BASMA SULAIMAN GHANDOURAH Department of Orthodontics and Pediatric Dentistry,
School of Dentistry, University of Michigan, Ann Arbor, MI, USA
MAIKO SUZUKI Division of Biosciences, Ohio State University, College of Dentistry,
Columbus, OH, USA
JULIEN TARABEUX Hôpitaux Universitaires de Strasbourg, Laboratoires de diagnostic géné
LIVIA M. A. TENUTA Department of Cariology, Restorative Sciences and Endodontics, School
of Dentistry, University of Michigan, Ann Arbor, MI, USA
xviii Contributors
KOSTAS VERDELIS Department of Restorative Dentistry/Comprehensive Care, School of

Dental Medicine, University of Pittsburgh, Pittsburgh, PA, USA; Department of
Endodontics and Center for Craniofacial Regeneration, School of Dental Medicine,
University of Pittsburgh, Pittsburgh, PA, USA
XIAOFANG WANG Department of Biomedical Sciences and Center for Craniofacial Research
and Diagnosis, Texas A&M University College of Dentistry, Dallas, TX, USA
YUANYUAN WANG Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
DI WU Department of Biostatistics, Gillings School of Global Public Health, University of
North Carolina-Chapel Hill, Chapel Hill, NC, USA; Department of Periodontology,
UNC School of Dentistry, University of North Carolina-Chapel Hill, Chapel Hill, NC,
USA
JINGYI WU Department of Biomedical Sciences and Center for Craniofacial Research and
Diagnosis, Texas A&M University College of Dentistry, Dallas, TX, USA; Stomatological
Hospital, Southern Medical University, Guangzhou, Guangdong, P.R. China
WEN’AN XU Department of Developmental Dentistry, School of Dentistry, The University of
Texas Health Science at San Antonio, San Antonio, TX, USA
YASUO YAMAKOSHI Department of Biochemistry and Molecular Biology, School of Dental
Medicine, Tsurumi University, Yokohama, Japan
XU YANG Department of Oral Biology, School of Dental Medicine, University of Pittsburgh,
Pittsburgh, PA, USA
PAMELA C. YELICK Department of Orthodontics, Tufts University School of Dental Medicine,
Boston, MA, USA; Department of Biomedical Engineering, Tufts University, Boston, MA,
USA; Department of Cell, Molecular, and Developmental Biology, Sackler School of
Graduate Biomedical Sciences, Tufts University School of Medicine, Boston, MA, USA
L. K. ZAUGG Department of Craniofacial Development and Stem Cell Biology, Centre for
Craniofacial and Regenerative Biology (CCRB), Dental Institute, King’s College London,
London, UK
YOUBIN ZHANG Department of Oral Biology, College of Dentistry, The University of Illinois
at Chicago, Chicago, IL, USA
LI ZHENG Department of Orthodontics and Pediatric Dentistry, School of Dentistry,
University of Michigan, Ann Arbor, MI, USA; Department of Orthodontics, The Ohio State
University College of Dentistry, Columbus, OH, USA
Part I
Protocols for Establishment of Dental Cells Lines and Animal

Models
Chapter 1
Microdissection and Isolation of Mouse Dental Epithelial

Cells of Continuously Growing Mouse Incisors
Hidemitsu Harada and Keishi Otsu
Abstract
Mouse incisors are regenerative tissues, which grow continuously throughout life and are good model for
the study of epithelial stem cells. The study of dental epithelial stem cells allows investigation of a variety of
basic biological processes in the context of the stem cells. The ability to analyze dental epithelial stem cells
in vitro has emerged as a powerful tool to understand how teeth are constructed and the signaling pathways
that regulate ameloblast developmental processes. Here, we describe in detail our protocols for the culture
of dental epithelial stem cells and the production of the cell lines. These techniques allow us to reproduce
the differentiation process of ameloblasts and estimate the effect of specific genes ex vivo, as well as are a tool
for studies on the mechanisms of normal and abnormal amelogenesis. They may also be applied to studies
on other aspects of developmental biology and regenerative medicine using stem cells.
Key words Tooth development, Cell culture, Dental epithelial stem cell
1 Introduction
Tooth development proceeds with sequential and reciprocal epithe-

lial-mesenchymal interactions, which are regulated by a number of
soluble proteins [1–3]. To date, most of the studies on tooth
development have been done with conventional histological meth-
ods and organ culture. By those methods, however, it is very hard
to analyze the molecular dynamics and intra- and extracellular
signaling in the process of cell differentiation and maintenance of
stem cells.
Dental epithelial stem cell culture allows us to overcome this
problem to a great extent. The molecular signals of dental epithelial
stem cell differentiation remain comparable to the in vivo tooth
development in terms of function, which attributes make them
more suitable than histological studies. Moreover, the role of spe-
cific growth factors and genes under their influence can be assessed
by using a variety of culture condition [4–6].
Petros Papagerakis (ed.), Odontogenesis: Methods and Protocols, Methods in Molecular Biology, vol. 1922,
https://doi.org/10.1007/978-1-4939-9012-2_1, © Springer Science+Business Media, LLC, part of Springer Nature 2019
3
4 Hidemitsu Harada and Keishi Otsu
Rodent incisors are known to be continuously growing teeth

that are maintained by both the cell proliferation at the apical end
and the attrition of the incisal edge. It is good model to study the
molecular mechanisms of asymmetric and slow-cycling cell division
especially of stem cell [5]. This type of tooth had a special epithelial
structure for the maintenance of stem cells, showing the bulbous
epithelial protrusion at the apical end [5].
Here, we have established dental epithelial stem cell culture
system, which opens the door to the direct intracellular signal
analysis of the developing tooth. Establishment of immortalized
dental epithelial stem cell lines provides stable experimental condi-
tion. A representative application of this method is useful for the
examination of dental epithelial stem cells from genetically mod-
ified mice. Further, we can examine the differentiation processes of
ameloblasts from undifferentiated epithelial cells in the presence of
and absence of dental mesenchymal cells [7].
Together, these two techniques provide information on the
growth, differentiation, and development of tooth germs and on
the influences of various factors on these processes. The results
obtained by the use of these methods usually provide insight into
the mechanisms of maintenance of stem cells and ameloblast differ-
entiation process.
In this chapter, we will describe our protocols for dental epi-
thelial stem cell culture and establishment of immortalized dental
epithelial stem cells, as well as give some technical tips.
2 Materials
2.1 Incisor All materials should be sterile.

Extraction and Culture
1. Scissors.
2. Fine forceps (Dumont, #5).
3. Culture dishes (35 mm).
4. Culture dishes coated with fibronectin (35 mm).
5. Nonadhesive culture dish (35 mm).
6. Disposable plastic syringes (1 cc).
7. Needles (18G, 25G).
8. Stereomicroscope with underneath transmitted light (Fig. 1).
9. DMEM/F12.
10. Penicillin/streptomycin.
11. Fibroblast growth factor 2 (Fgf2).
12. Epidermal growth factor (Egf).
13. B27 supplement (Invitrogen).
14. Collagenase.
Isolation of Dental Stem Cells 5
Fig. 1 Stereomicroscope for the separation of incisors from mandible. Overall

view of the system
2.2 Media 1. Working medium: DMEM/F12 supplemented with 50 U/mL

penicillin/streptomycin.
2. Culture medium: DMEM/F12 supplemented with Fgf2
(25 ng/mL), EGF (100 ng/mL), and B27 supplement (2%).
3. Collagenase solution: 2% collagenase (weight/volume) in
working medium.
2.3 General 1. Humidified incubator at 37 C and 5% CO2.

Equipment 2. Tissue culture hood.
3. 37 C water bath.
4. Stereomicroscope with underneath transmitted light (Fig. 1).
3 Methods
All procedures must be performed under sterile techniques with

great attention. Use clean and detergent-free glassware. All of
the following steps should be performed on ice or at 4 C. All
procedure is shown in Fig. 2.
3.1 Extraction of an 1. Rinse surgical equipment in 70% ethanol prior to use.

Incisor Germ from a 2. Place the mouse (postnatal days 3–7) in 100 mm dishes con-
Mandible taining working medium on ice (see Note 1).
3. Wash several times in ice-cold working medium.
separate
Lower incisor 2%collagenase
From PN3-7d mice 4˚C, 12h
Cut off
dental papilla cells

of apical end
Labial cervical loop epithelium
culture
Fig. 2 Protocol for the separation of labial cervical loop of a mouse incisor
4. Use forceps and scissors to remove the mouse heads.

5. Remove a lower jaw from a head while observing under a
stereomicroscope.
6. Use the tip of an 18-G needle or fine forceps to remove tooth
germs of incisors from lower jaws.
7. Use needles to remove excess non-dental tissue (see Note 2).
8. Use fine forceps to remove dental follicle around the apical end
of an incisor (Fig. 3).
9. Transfer incisor germs to fresh working medium, and keep
them on ice.
3.2 Separation of a 1. Use forceps to transfer incisor germs in the medium including
Dental Epithelial Cell 2% collagenase in the culture dishes.
Sheet from an 2. Incubate it at 4 C for 3–12 h or 37 C for 30 min–1 h.
Incisor Germ 3. Note the laxation between a dental epithelial sheet and an
incisor germ (Fig. 4) (see Note 3).
4. Use forceps to pinch the incisal tip of labial epithelial cell sheet
at the surface of the enamel and then to separate from incisor
germ slowly (Fig. 5) (see Note 4).
5. Transfer a labial epithelial cell sheet to fresh working medium,
and keep them on ice (Fig. 6).
3.3 Culture of a 1. Use an 18-G needle or surgical knife to divorce a labial cervical
Labial Cervical Loop loop epithelium from a labial dental epithelial cell sheet
Epithelium (Fig. 7).
2. Pipette a labial cervical loop epithelium, and transfer it to
fibronectin-coated culture dish (35 mm).
3. The labial cervical loop epithelium binds the surface of the dish,
and the cells proliferate and expand around the colony (Fig. 8).
Fig. 3 Process of separation of dental follicle from a mouse incisor. Arrows indicates the dental follicle
Fig. 4 Appearance of an incisor germ after collagenase treatment. Arrow

indicates a labial cervical loop epithelium
Fig. 5 Process of separation of a labial epithelial cell sheet
Fig. 6 Appearance of a labial cervical loop. An arrow indicates a labial cervical

loop epithelium
3.4 Immortalizing of 1. The cells from a labial cervical loop epithelium expand on the
Dental Epithelial Stem fibronectin-coated dish and proliferate slowly at first (see Note 5).
Cells 2. When the diameter of the colony becomes about 1–2 cm, it can
be passaged.
Fig. 7 Cutoff between labial cervical loop and inner enamel epithelium. (a) Appearance of labial epithelial cell
sheet before cutting off. (b) Appearance of labial epithelial cell sheet after cutting off. Arrows indicates a labial
cervical loop epithelium
Fig. 8 Culture of labial cervical loop epithelial cells. (a) Appearance of a labial cervical loop epithelium that
binds to fibronectin-coated culture dish. (b) Appearance of expanding labial cervical loop cells
3. After the passage several times, the cells become to proliferate

actively.
4. 500–5000 cells are seeded on the 10 cm culture dish, and a
small number of colonies come in. One of the colonies is peeled
off using a cloning cylinder and is seeded on 10 cm culture dish
again. The treatment is repeated two to three times. Continue
to culture the rest of the cells for backup preservation.
5. Finally, carry out a single-cell cloning by limited dilution cul-
ture method.
6. After the establishment of immortalized dental epithelial stem
cells from the colony, confirm that all cells are cytokeratin
14 positive by immunostaining or FACS.
7. When the cells are seeded on nonadhesive culture dish, the
immortalized dental epithelial stem cells become to be floating
sphere in the culture medium (Fig. 9) [8].
Fig. 9 Appearance of sphere of immortalized dental epithelial stem cells on

nonadhesive culture dish
4 Notes
1. Mouse incisors of postnatal days 3–7 are the most appropriate

for the experiment. Before postnatal day 7 is also suitable for
extraction of tooth germs, since bone calcification is not com-
plete at this stage. Forceps and needles can easily remove bone
and other tissues. Additionally, before postnatal day 3, it is
difficult to remove the bone from the apical end of mouse
incisors, since a labial cervical loop epithelium connects the
surrounding connective tissues tightly.
2. Removing the excess of non-dental tissue and dental follicle
facilitates the observation of the epithelial layer and a labial
cervical epithelium.
3. Be careful not to treat it excessively under collagenase solution.
Excess treatment complicates a discrimination between a labial
epithelial sheet and connective tissue around an incisor germ.
4. When removing the labial epithelial cell sheet from incisor
germ, be very careful not to break off the labial epithelial cell
sheet (ameloblasts) and a labial cervical epithelium from the
labial epithelial cell sheet.
5. Basically, mouse dental epithelial stem cells can immortalize
spontaneously and easily without the transfection of virus
gene and/or some gene in the presence of Fgf2 and EGF.
Acknowledgments
This work was supported, in part, by KAKENHI (26670805 to

HH) from the Ministry of Education, Culture, Sports, Science, and
Technology of Japan. All authors state that they have no conflicts of
interest.
References
1. Thesleff I, Mikkola M (2002) The role of putative stem cells in dental epithelium and
growth factors in tooth development. Int Rev their association with Notch and FGF signaling.
Cytol 217:93–135 J Cell Biol 147:105–120
2. Thesleff I, Keranen S, Jernvall J (2001) Enamel 6. Harada H, Ichimori Y, Yokohama-Tamaki T,
knots as signaling centers linking tooth morpho- Ohshima H, Kawano S, Katsube K, Wakisaka S
genesis and odontoblast differentiation. Adv (2006) Stratum intermedium lineage diverges
Dent Res 15:14–18 from ameloblast lineage via Notch signaling.
3. Jernvall J, Thesleff I (2000) Reiterative signaling Biochem Biophys Res Commun 340:611–616
and patterning during mammalian tooth mor- 7. Morotomi T, Kawano S, Toyono T, Kitamura C,
phogenesis. Mech Dev 92:19–29 Terashita M, Uchida T, Toyoshima K, Harada H
4. Yokohama-Tamaki T, Ohshima H, Fujiwara N, (2005) In vitro differentiation of dental epithe-
Takada Y, Ichimori Y, Wakisaka S, Ohuchi H, lial progenitor cells through epithelial-
Harada H (2006) Cessation of Fgf10 signaling, mesenchymal interactions. Arch Oral Biol
resulting in a defective dental epithelial stem cell 50:695–705
compartment, leads to the transition from 8. Chavez MG, Yu W, Biehs B, Harada H, Snead
crown to root formation. Development ML, Klein OD (2013) Characterization of den-
133:1359–1366 tal epithelial stem cells from the mouse incisor
5. Harada H, Kettunen P, Jung HS, Mustonen T, with 2D and 3D platforms. Tissue Eng Part C
Wang YA, Thesleff I (1999) Localization of Methods 19(1):15–24
Chapter 2
Establishment of an Immortalized Mouse Bmp2 Knockout

Dental Papilla Mesenchymal Cell Line
Wen’an Xu and Shuo Chen
Abstract
Bone morphogenetic protein 2 (Bmp2) is essential for dentin formation. Bmp2 cKO mice exhibited similar
phenotype to dentinogenesis imperfecta (DGI), showing dental pulp exposure, hypomineralized dentin,
and delayed odontoblast differentiation. As it is relatively difficult to obtain primary Bmp2 cKO dental
papilla mesenchymal cells and to maintain a long-term culture of these primary cells, availability of
immortalized deleted Bmp2 dental papilla mesenchymal cells is critical for studying the underlying mecha-
nism of Bmp2 signal in odontogenesis. Here we describe the generation of an immortalized deleted Bmp2
dental papilla mesenchymal (iBmp2ko/ko-dp) cell line by introducing Cre fluorescent protein (GFP) into
the immortalized mouse floxed Bmp2 dental papilla mesenchymal (iBmp2flox/flox-dp) cells.
Key words Bone morphogenetic protein 2, Dental papilla mesenchymal cell line, Bmp2 floxed mice,
Knockout, Dentin formation, Dentinogenesis imperfecta
1 Introduction
Tooth development involves sequential and reciprocal interactions

between dental epithelial and mesenchymal cells and proceeds
through a series of cytodifferentiations in specific spatial-temporal
patterns [1]. Dentinogenesis is a complex process in which multiple
signaling pathways converge to induce dentin formation and is
controlled by many growth and transcription factors [2]. The
bone morphogenetic proteins (Bmps) are structurally related to
the transforming growth factor beta (TGF-β) superfamily and
were originally identified by their capacity to induce ectopic bone
formation in rodents [3, 4]. Among the Bmp family members,
Bmp2 expression is observed in dental cells during tooth develop-
ment and formation [5]. Also, Bmp2 promotes dental pulp stem
cell commitment to odontoblast lineages [6] and induces dental
pulp cell differentiation [7, 8]. Bmp2 conditional knockout (cKO)
mice displayed abnormal tooth phenotypes with delayed
13
14 Wen’an Xu and Shuo Chen
odontoblast differentiation, abnormal dentin tubules, and

decreased tooth-related gene expression [9–11].
However, detail understandings of the molecular mechanisms
of Bmp2 exerting its effects on tooth development and formation
remain elusive in particular during postnatal tooth development as
homozygous mutant embryos for Bmp2 showed developmental
abnormalities and died at embryo day 9.5 [12]. Recently, condi-
tional Bmp2 knockout (cBmp2-KO) mice have been generated and
revealed important roles of Bmp2 in late stages of organogenesis
including the bone, heart, and uterus [13–18].
Unlike the bone and other tissues, it is relatively hard to collect
enough amounts of primary dental papilla mesenchymal cells from
a mouse single tooth. Therefore, generation of a Bmp2 KO dental
papilla mesenchymal cell line would be a valuable tool for studying
the effects of Bmp2 on dental cell lineages.
2 Materials
2.1 Vectors for A 10-kb SpeI gene fragment that contains exon 2 and exon 3 was
Conditional Gene subcloned into pBluescript. One LoxP site followed by phospho-
Targeting glycerol kinase neomycin resistance (Pgk-Neo) with two flanking
sites of expression cassette was blunt cloned into Avr2 site that is in
the intron downstream of Bmp2 exon 3. Another LoxP site was
inserted into the XhoI site that is upstream of exon 3. The 50 end of
the targeting vector was constructed by cloning 6 kb of Bmp2
homologous sequence containing exon 2 that has putative initiator
methionine and intron 2 into SalI and XhoI sites.
2.2 Animals C57BL/6 mice.
2.3 Cells and Cell Embryonic stem (ES) cells of mouse.

Culture Phosphate-buffered saline (PBS), pH 7.4.
Cell culture 6-well plates.
Alpha minimum essential medium (α-MEM).
G418-containing medium.
Digestion enzyme solution: 3 mg/mL collagenase type I and
4 mg/mL dispase II.
50 mL centrifuge tube.
Antibiotics: 100 U/mL penicillin and 400 μg/mL streptomycin.
2.4 Virus for Lentivirus carrying the SV 40 large T antigen gene.

Infection Adenovirus with Cre recombinase and green fluorescent protein
(GFP).
Immortalized Bmp2 KO Dental Papilla Mesenchymal Cells 15
2.5 Western Blot RIPA buffer: 1 PBS, 1% Nonidet P-40, 0.5% sodium deoxycho-
late, 0.1% SDS, 10 mg/mL phenylmethylsulfonyl fluoride
(PMSF), 50 KIU/mL aprotinin, and 100 mM sodium
orthovanadate.
TBST buffer: 10 mM Tris–HCl, pH 7.5, 100 mM NaCl, and 0.1%
Tween-20.
12% SDS-PAGE gel.
Trans-Blot membranes.
5% nonfat milk.
2.6 Antibodies Primary anti-SV40 large T antigen monoclonal antibody.

Secondary antibody with Alexa Fluor® 568 red fluorescent labeling.
Rabbit polyclonal anti-mouse Bmp2 antibody.
Horseradish peroxidase-conjugated anti-rabbit IgG.
Mouse monoclonal anti-BrdU antibody.
Antibodies (goat-anti-mouse) with Alexa Fluo1® 488 green.
2.7 Other Materials Stereomicroscope.

Goat serum.
5-Bromo-20 -deoxyuridine (BrdU).
Acetone.
Methanol.
3 Methods
3.1 Generation of A conditional allele of the mouse Bmp2 gene was created by intro-
Bmp2 Floxed Mice ducing Cre recombinase recognition sites (loxP), which were
placed upstream and downstream of exon 3 to excise the protein-
coding region in exon 3 of the Bmp2 gene [19].
1. The embryonic stem (ES) cells were transfected with a linear-
ized targeting vector by electroporation and selected in G418-
containing medium as described previously [20]. The geno-
types of selected clones were analyzed by Southern blot using
50 external probe (SalI-SalI fragment, 0.7 kb) with SpeI-
digested DNA.
2. The positive clones were microinjected into blastocysts derived
from C57BL/6 mice. The chimeras were bred to C57BL/6
females, and F1 agouti offspring were analyzed by polymerase
chain reaction (PCR) for the presence of Bmp2 floxed allele.
3.2 Establishment of 1. The dental papilla mesenchyme was separated with enamel
Immortalized Floxed tissues from the first molars of 1-day-old floxed Bmp2 mice
Bmp2 Dental Papilla under stereomicroscope with tweezers. The dental papilla mes-
Mesenchymal Cells enchyme was washed with phosphate-buffered saline (PBS)
and digested for 1 h at 37 C in a solution of 3 mg/mL
collagenase type I and 4 mg/mL of dispase (see Note 1).
2. Primary mouse papilla mesenchymal cells in passage 3 were
grown about 85% confluence and infected by lentivirus carry-
ing the SV 40 large T antigen gene following the manufac-
turer’s protocol.
3. Two days after infection, the primary cells were replated at a
low density to get separated colonies. Several colonies were
formed, and well-isolated colonies were removed selectively
and replated at low densities to obtain the secondary selection.
4. Several single cells which grew were expanded into cell lines
and passaged at least 30–50 times over a 5–12-month period
(see Note 2).
5. Genomic DNAs were isolated from immortalized floxed papilla
mesenchymal (iBmp2flox/flox-dp) cells of passage 50 and the
primary cells of passage 3 for detection of transformation (see
Note 3).
6. The iBmp2flox/flox-dp and primary cells were seeded on cover-
slips in a 6-well plate and cultured for 48 h in standard α-MEM
medium. The coverslips were rinsed with PBS and fixed with
cold acetone and methanol (1:1). The cells were blocked with
10% goat serum and incubated with a primary anti-SV40 large
T antigen monoclonal antibody for 2 h at 37 C. Then the cells
were washed 3 for 5 min with 1 PBS and incubated with the
secondary antibody with Alexa Fluor® 568 red fluorescent
labeling for 1 h at room temperature. Microphotograph was
obtained under a Nikon microscope using a Nikon Cool pix
4500 digital camera (see Note 4).
7. Morphology of the iBmp2flox/flox-dp and primary dental papilla
mesenchymal cells was observed by a light inverted
microscope.
8. The iBmp2flox/flox-dp and primary cell proliferation was iden-
tified by 5-bromo-20 -deoxyuridine (BrdU) incorporation.
Images were obtained in a Nikon inverted microscope, and
proliferative cells were expressed as a percentage of the number
of BrdU-positive cells relative to the total number of Hoechst-
positive nuclei (see Note 5).
3.3 Generation of 1. Adenovirus with Cre recombinase and green fluorescent pro-
Immortalized Bmp2 KO tein (GFP) was obtained from Vector Biolabs and added to the
Dental Papilla iBmp2flox/flox-dp cells. The cells were transduced overnight for
Mesenchymal Cells 14 h and then recovered in cultured medium.
2. GFP-positive cells were observed under a Nikon inverted fluo-

rescent microscope. The positive cells were selectively picked
up with a pipette and replated at low densities to obtain further
cell growth.
3. Genomic DNAs were isolated from the iBmp2flox/flox-dp and
immortalized mouse Bmp2 KO dental papilla mesenchymal
(iBmp2ko/ko-dp) cells using DNA purification kit (see Note 6).
4. Proteins were isolated from the iBmp2flox/flox-dp
and iBmp2ko/ko-dp cells. Bmp2 protein expression in the
iBmp2flox/flox-dp and iBmp2ko/ko-dp cells was detected by
Western blot assay using anti-Bmp2 antibody (see Note 7).
4 Notes
1. The dental papilla mesenchymal cells were grown with alpha

minimum essential medium containing 10% fetal calf serum
plus penicillin (100 U/mL) and streptomycin (100 μg/mL)
and cultured at 37 C in a humidified atmosphere of air con-
taining 5% CO2. The medium was refreshed every 2 days, and
cells were spread after reaching confluence.
2. Of the several secondary cell lines established, line
iBmp2flox/flox-dp (immortalized floxed Bmp2 dental papilla
mesenchymal) of passage 50 and primary floxed Bmp2 dental
papilla mesenchymal cells of passage 3 were used for the fol-
lowing characterization.
3. Two hundred nanograms of DNA was diluted in a 25 μL PCR
mix of 1 PCR buffer containing 10 ƿmoL of forward and
reverse primers, 1 U Red Taq DNA polymerase, and 2.5 mM
dNTPs (Sigma-Aldrich). Simian virus 40 sequences were
accessed in GenBank and specific primers:
Forward: 5-AGCAGACACTCTATGCCTGTGTGGAGTAA
G-30
Reversed: 5-GACTTTGGAGGCTTCTGGATGCAACTGA
G-30
PCR conditions: 4 min at 94 , 35 cycles of 1 min at 94 C,
1 min at 62 C, and 2 min at 72 C, followed by 10 min at
72 C. The amplified products were run on a 1%
agarose gel.
4. For negative control, the primary SV40 antibody was replaced
by mouse IgG I. For cell nucleus staining, the cells were treated
with Hoechst. Images of Alexa Fluor® 568 staining of the SV40
protein were obtained at the Core Optical Imaging Facility
under the same parameters in a Nikon inverted microscope.
5. iBmp2flox/flox-dp and primary cells were transferred into 6-well
glass slides and incubated with 30 μM BrdU in culture medium
for 4 h. The cells were treated with a mouse monoclonal anti-

BrdU antibody, followed by a 1:1000 dilution of the antibodies
(goat-anti-mouse) with Alexa Fluor® 488 green. For nucleus
staining, the cells were incubated with a 1:5000 dilution of
Hoechst.
6. PCR genotyping was performed by amplification of the
floxed/floxed (Bmp2 flox/flox) and recombinant (Bmp2ko/ko)
alleles using two pair primers:
Bmp2flox/flox, forward: 50 -GATGATGAGGTTCTTGGCGG-30
Reversed: 50 -AGGGTTTCAGGTCAGTTTCCG-30
Bmp2ko/ko, forward: 50 -GATGATGAGGTTCTTGGCGG-30
Reversed: 50 -AGCATGAACCCTCATGTGTTGG-30
7. The iBmp2flox/flox-dp and iBmp2ko/ko-dp cells were then
washed with 1 cold PBS and lysed with RIPA buffer (1 PBS,
1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS,
10 mg/ml phenylmethylsulfonyl fluoride (PMSF),
50 KIU/mL aprotinin, 100 mM sodium orthovanadate).
Whole cell lysates were resolved by a 12% SDS-PAGE gel and
transferred to Trans-Blot membranes. The membranes were
blocked with 5% nonfat milk in TBST buffer (10 mM
Tris–HCl, pH 7.5, 100 mM NaCl, 0.1% Tween-20) for
60 min at room temperature. After washing, the membranes
were incubated with rabbit polyclonal anti-mouse Bmp2 anti-
body with appropriate dilution (1:500) overnight at 4 C. After
washing, the membrane was incubated with the secondary
antibody (horseradish peroxidase-conjugated anti-rabbit IgG)
at room temperature for 60 min. Immunoreactivity was deter-
mined using ECL chemiluminescence reagent.
Acknowledgments
We are grateful to the core facility center at the University of Texas

Health Center at San Antonio, Texas, which performed cell cycle
experiments. This research was supported by the National Insti-
tutes of Health (NIH), the National Institute of Dental and Cra-
niofacial Research (NIDCR, DE019802), and partially the Natural
Science Foundation of China (81170929).
References
1. Linde A, Goldberg M (1993) Dentinogenesis. 3. Urist MR (1965) Bone: formation by autoin-

Cri Rev in Oral Biol Med 4(5):679–728 duction. Science 150(3698):893–899
2. Thesleff I (2003) Epithelial-mesenchymal sig- 4. Wozney JM, Rosen V, Celeste AJ et al (1988)
nalling regulating tooth morphogenesis. J Cell Novel regulators of bone formation: molecular
Sci 116(9):1647–1648 clones and activities. Science 242
(4885):1528–1535
5. Åberg T, Wozney J, Thesleff I (1997) Expres- 13. Bandyopadhyay A, Tsuji K, Cox K et al (2006)
sion patterns of bone morphogenetic proteins Genetic analysis of the roles of BMP2, BMP4,
(Bmps) in the developing mouse tooth suggest and BMP7 in limb patterning and skeletogen-
roles in morphogenesis and cell differentiation. esis. PLoS Genet 2:e216
Dev Dyn 210(4):383–396 14. Tsuji K, Bandyopadhyay A, Harfe BD et al
6. Yang X, Van Der Kraan PM, Bian Z et al (2009) (2006) BMP2 activity, although dispensable
Mineralized tissue formation by BMP2- transfor bone formation, is required for the initiation
fected pulp stem cells. J Dent Res 88 of fracture healing. Nat Genet 38:1424–1429
(11):1020–1025 15. Ma L, Lu MF, Schwartz RJ et al (2005) Bmp2
7. Chen S, Gluhak-Heinrich J, Martinez M et al is essential for cardiac cushion epithelial-
(2008) Bone morphogenetic protein 2 med- mesenchymal transition and myocardial pat-
iates dentin sialophosphoprotein expression terning. Development 132:5601–5611
and odontoblast differentiation via NF-Y sig- 16. Rivera-Feliciano J, Tabin CJ (2006) Bmp2
naling. J Biol Chem 283:19359–19370 instructs cardiac progenitors to form the
8. Cho YD, Yoon WJ, Woo KM et al (2010) The heart-valve- inducing field. Dev Biol
canonical BMP signaling pathway plays a cru- 295:580–588
cial part in stimulation of dentin sialophospho- 17. Lee KY, Jeong JW, Wang J et al (2007) Bmp2 is
protein expression by BMP-2. J Biol Chem critical for the murine uterine decidual
285:36369–36376 response. Mol Cell Biol 27:5468–5478
9. Feng J, Yang G, Yuan G et al (2011) Abnorm- 18. Singh AP, Castranio T, Scott G et al (2008)
alities in the enamel in bmp2-deficient mice. Influences of reduced expression of maternal
Cells Tissues Organs 194:216–221 bone morphogenetic protein 2 on mouse
10. Yang W, Harris MA, Cui Y et al (2012) Bmp2 is embryonic development. Sex Dev 3:134–141
required for odontoblast differentiation and 19. Ma L, Martin JF (2005) Generation of a Bmp2
pulp vasculogenesis. J Dent Res 91:58–64 conditional null allele. Genesis 42:203–206
11. Guo F, Feng J, Wang F et al (2014) Bmp2 20. Lu MF, Cheng HT, Kern MJ et al (1999) prx-1
deletion causes an amelogenesis imperfecta functions cooperatively with another paired-
phenotype via regulating enamel gene expres- related homeobox gene, prx-2, to maintain
sion. J Cell Physiol 230:1871–1882 cell fates within the craniofacial mesenchyme.
12. Zhang H, Bradley A (1996) Mice deficient for Development 126:495–504
BMP2 are nonviable and have defects in
amnion/chorion and cardiac development.
Development 122:2977–2986
Chapter 3
Establishment of Stable Cell Lines from Primary Human

Dental Pulp Stem Cells
Elizabeth Guirado, Youbin Zhang, and Anne George
Abstract
This protocol is for the isolation of primary human dental pulp stem cells (DPSCs) from adult extracted
molars and for the generation of high-titer lentivirus for in vitro infection of the DPSCs. Stable cell lines of
dental pulp stem cells are generated, maintained in culture, and used for subsequent experiments.
Key words Lentivirus, Gene transfer techniques, Genetic transduction, Genetic recombination,
Somatic stem cells
1 Introduction
Lentiviral expression vector systems have been used extensively to

overexpress proteins in mammalian cells. This method of gene
transfer is used for otherwise difficult to transfect target cells and
has allowed the study of protein function in the same [1, 2]. Briefly,
a gene of interest is cloned into a commercially available,
replication-incompetent lentiviral vector. The remaining genomic
material necessary for virion assembly is co-transfected into pack-
aging cell lines, commonly human embryonic kidney cells
293 (HEK 293 cells), which will produce infectious transgenic
lentiviruses in culture. This transgenic virus is then used to infect
and deliver the gene of interest into a target cell type [1–4]. Because
lentiviruses naturally integrate into the host genome, the gene of
interest is now stably passed down to daughter cells.
Many different lentiviral transfer plasmids exist into which one
can clone their sequence of interest. In this protocol, a lentiviral
transfer plasmid is used that contains a CMV promoter-driven gene
of interest, an SV40 promoter-driven GFP reporter, and a puromy-
cin resistance selection marker for mammalian cells. Dental pulp
stem cells (DPSCs) are used in these studies for their ability to
differentiate into dentin-forming odontoblasts. Understanding
21
22 Elizabeth Guirado et al.
the molecular pathways by which dental pulp stem cells differenti-

ate into dentin-secreting odontoblasts promises an opportunity to
use these stem cells for regenerative therapies of injured dentin
[5–7].
2 Materials
2.1 Isolation of 1. Adult molars.

Primary Human DPSCs 2. Povidone-iodine solution.
3. Sterilized dental fissure burs.
4. Type I collagenase (3 mg/mL).
5. Dispase (4 mg/mL).
6. 70 μm strainer.
7. Dulbecco’s Modified Eagle Medium: 4 g/L D-glucose, 4 mM
L-glutamine, 1 mM sodium pyruvate, and phenol red.
8. Antibiotic-antimycotic 100: 10,000 units/mL of penicillin,

10,000 μg/mL of streptomycin, and 25 μg/mL of Gibco
amphotericin B.
9. 0.05% trypsin-EDTA and phenol red.
10. Dimethyl sulfoxide (DMSO).
11. Defined fetal bovine serum (HyClone).
12. Mr. Frosty™ freezing container (Thermo Scientific™).
2.1.1 Cell Maintenance: 1. 500 mL Dulbecco’s Modified Eagle Medium: 4 g/L D-glu-
DMEM Media (per 500 mL) cose, 4 mM L-glutamine, 1 mM sodium pyruvate, and
phenol red.
2. 50 mL defined fetal bovine serum (HyClone).
3. 5 mL antibiotic-antimycotic 100 (1% w/v).
2.2 Establishment of 1. 293FT cells.

Stable Cell Lines from 2. 100 mm, 150 mm tissue culture dishes.
Primary Human DPSCs
3. 6-well plate.
4. Ultracentrifuge tubes, 38.5 mL and 25 89 mm.
5. Dulbecco’s Modified Eagle Medium: 4 g/L D-glucose, 4 mM
L-glutamine, 1 mM sodium pyruvate, and phenol red.
6. Antibiotic-antimycotic 100: 10,000 units/mL of penicillin,

10,000 μg/mL of streptomycin, and 25 μg/mL of Gibco
amphotericin B.
7. Phosphate-buffered saline without Ca2+ and Mg2+.
8. Defined fetal bovine serum (HyClone).
9. Sodium phosphate, dibasic.
Stable Cell Lines From Primary Human DPSCs 23
10. Sodium chloride.

11. Sodium hydroxide.
12. Tris–Cl.
13. Potassium chloride.
14. Glycerol.
15. Cell Culture Grade Water, Deionized, UltraPure, Endotoxin-
Free, Sterile.
16. 0.45 μm low protein-binding filter flask (Millipore).
17. Puromycin dihydrochloride (10 mg/mL in H2O).
2.3 Transfection 1. 1 mM Tris-Cl, pH 7.05.

Reagents 2. Calcium chloride 2 M solution.
3. 15% glycerol PBS solution (per 50 mL).
4. 2 HBS buffer (per 500 mL): 280 mM NaCl, 10 mM KCl,
1.5 mM Na2HPO42H2O, 12 mM dextrose, and 50 mM
HEPES (see Note 1).
3 Methods
3.1 Isolation of 1. Third molars were collected from adult patients, decontami-
Primary Human DPSCs nated with povidone-iodine solution. Teeth were sectioned
longitudinally using sterilized dental burs to reveal the pulp
chamber. Exposed pulp tissues were gently separated from the
crown and root, collected, and enzymatically digested with
type I collagenase (3 mg/mL) and dispase (4 mg/mL) for
1 h at 37 C. Single-cell suspensions were obtained by passing
the cells through a 70 μm strainer [8].
2. Cells were counted and seeded at a density of 1.8 104/cm2.
Cell cultures were maintained with Dulbecco’s Modified Eagle
Medium supplemented with 10% fetal bovine serum, 1%
antibiotic-antimycotic 100, at 37 C with 5% CO2. The
medium was refreshed the next day after initial cell attachment
and thereafter at three times per week. Cells exhibit a fibroblast-
like morphology when observed under the microscope.
3. Cells were detached by trypsinization whenever 80–90% con-
fluent using 0.05% trypsin-EDTA and phenol red solution and
were replated at the same density. Colony-forming units (aggre-
gates of 50 cells) derived from dental pulp tissue averaged
22–70 colonies/104 cells plated, as previously published [8].
4. For storage, cells were trypsinized, centrifuged at 218 g for
5 min, and resuspended in 90% FBS-10% dimethyl sulfoxide
(DMSO). Resuspension was frozen at 80 C (see Note 2).
Cells were transferred to liquid nitrogen within a week.
3.2 Establishment of Day 0:

Stable Cell Lines from
1. Split four 100 mm dishes of 95% confluent 293FT cells into five
Primary Human DPSCs
to eight 150 mm dishes. For each dish, use 25 mL of DMEM
media (see Note 3).
2. Rock the plate gently to evenly distribute the cells.
3. Incubate the dishes at 37 C overnight. The cells should reach
90% confluence in 24 h.
Day 1:
1. Warm 330 mL of DMEM media to room temperature, and
replace cell’s old media.
2. Wait at least 4–6 h to start the transfection.
3. In a 50 mL conical tube, prepare the following mixture:
(a) 840 5–8 μL of 1 mM Tris-Cl
(b) 11.1 5–8 μg of lentivirus plasmid (e.g., pLenti-
hDMP1-GFP-2A-Puro or pLenti-GFP-2A-Puro)
(c) 10.7 5–8 μg of psPAX2 (Addgene)
(d) 5.8 5–8 μg of pMD2.G (Addgene)
(e) 5.1 5–8 μg of pHPV17
4. At this point, mix thoroughly, using bubbling to mix. Never
vortex the mixture (see Note 4).
(a) Add 120 5–8 μL of 2 M CaCl2 solution drop by drop to
DNA solution while bubbling the mixture.
(b) Mix thoroughly by bubbling.
5. Add the above mixed DNA-CaCl2 solution to 960 5–8 μL
of 2 HBS dropwise, slowly, while bubbling.
(a) Wait about 15–20 min. Then mix again, and pour directly
into the 150 mm dish of cells, 2 mL per dish of cells (see
Note 5).
6. Put the tissue culture dishes back into the incubator.
Day 2:
1. Warm DMEM media to room temperature.
2. Check the cells using a fluorescent microscope; 20–30% of the
cells should be GFP positive.
3. Fifteen to sixteen hours after initial transfection, remove the
transfection media from the dishes, leaving only a little bit of
media behind. Add 1–2 mL of 15% glycerol PBS solution to
dishes, and wait for 1 min. Then wash each dish twice with
25 mL of 1 PBS. Then add 25 mL of fresh DMEM media to
each dish.
4. Put the tissue culture dishes back into incubator for 24 h.
Day 3:
1. Twenty-four hours post transfection, collect the media (lenti-
virus medium), and check the cells. More than 50% cells should
be GFP positive by now.
2. Add 25 mL of DMEM media.
3. Put cells back into the incubator. Cells detach very easily at this
point, so be very gentle.
4. Split the target cells (DPSCs) to be infected into a 6-well plate;
be sure to have enough cells per well to reach 60% confluence
the next day. For stem cells, we use the lowest passage cells to
infect virus and establish cell lines.
Day 4:
1. Another 24 h later, collect the virus-containing supernatant
into four 50 mL conical tubes, and centrifuge all media for
10 min at 872 g.
2. To concentrate lentivirus by ultracentrifugation, divide the
filtered virus-containing supernatant among six ultracentrifuge
tubes.
3. Centrifuge in a Beckman SW-28 rotor for at least 2 h at
68,383 g, 4 C.
4. Gently carry the centrifuge tubes back to the tissue culture
hood, and pour out the supernatant. There should be a tiny
semitransparent pellet at the bottom of each centrifuge tube.
5. Dry the side of each tube with Kimwipes.
6. Add 500 μL of cold serum-free DMEM media to every tube,
and resuspend the pellet by swirling and gentle pipetting. Do
not pipet too much because it will degrade the virus.
7. After resuspending all virus pellets, pipet the virus solution into an
Eppendorf tube, and add polybrene, to reach 5–10 ug/mL final
concentration. Maintain at room temperature for 15–20 min.
8. Wash the cells to be infected twice with 1 PBS, and add the
virus-/polybrene-/serum-free DMEM medium to the 6-well
plate of cells, 1 mL/well. Incubate the cells in the biosafety
cabinet for 4 h. Then add 3 mL of the full nutrient medium to
each well.
Day 5:
1. Wash out the virus-containing medium with 1xPBS, twice, and
add fresh DMEM media to target cells in 6-well plate.
2. Twenty-four to forty-eight hours later, check the target cells,
and finally add puromycin 1–10 μg/mL to begin the selection.
It is recommended to use 1 μg/mL of puromycin to select
PDL, HMSC, and DPSC cells and 5–10 μg/mL to select
tumor cell lines.
4 Notes
1. To make 2 HEPES-buffered saline (HBS) solution, dissolve

1.6 g of NaCl, 0.074 g of KCl, 0.027 g of Na2HPO42H2O,
0.2 g of dextrose, and 1 g of HEPES in a total volume of 90 mL
of distilled H2O. Adjust the pH to 7.05 with 0.5 N NaOH, and
then adjust the volume to 100 mL with distilled H2O. Filter
with 0.22 μm filter. This solution is stable at room temperature
for 6 months.
2. A freezing container is used to freeze the cells at a rate of
1 C/min in the 80 C freezer. After 24 h, cells can be
transferred to a liquid nitrogen tank. To prevent thawing dur-
ing transport, dry ice may be placed inside an insulated con-
tainer, and cryovials containing cells may be placed inside. To
thaw cells, it is best to thaw quickly. Once again, cryovials can
be placed inside an insulated container containing dry ice.
Once in the cell culture area, hold the cryovial inside a 37 C
water bath, making sure the water line is well below the cap in
order to prevent infiltration. Remove cryovial just before all the
ice melts, and proceed to resuspend in media. Add cryovial
contents to 50 mL centrifuge tube, and add culture media
dropwise for about 5 mL to allow cells to adjust to the new
osmolality. Add the remaining media and plate into tissue
culture dish.
3. It is important to use low-passage 293FT cells for the produc-
tion of viruses. To make sure the cells are always in the fastest
growth phase, never let the cells grow to 100% confluence.
Prepare and freeze stocks and use a new vial whenever making
more virus.
4. The transfection reagents should never be vortexed. “Bub-
bling” refers to the use of two pipettes, simultaneously, to
mix the transfection reagents. Using one hand, a pipette will
be used to introduce air (“bubbles”) to the bottom of the tube.
Using the other hand, another pipette will be used to introduce
the DNA mixture, dropwise and very slowly, into the tube.
5. Be careful not to tilt the plates too much, as cells may detach
easily.
References
1. Miller AD (1990) Retrovirus packaging cells. 3. Mosimann C, Zon LI (2011) Chapter 10—
Hum Gene Ther 1:5–14 Advanced zebrafish transgenesis with Tol2 and
2. Marino MP, Luce MJ, Reiser J (2003) Small- to application for Cre/lox recombination experi-
large-scale production of lentivirus vectors. In: ments. In: Detrich HW, Westerfield M, Zon LI
Federico M (ed) Lentivirus gene engineering (eds) Methods in cell biology, vol 104. Academic
protocols. Methods in molecular biology, vol Press, Cambridge, MA, pp 173–194
vol. 229. Humana Press, New York
4. Austin J (2001) Transgenes. In: Brenner S, cells transplanted in the isogenic mouse spleen.
Miller JH (eds) Encyclopedia of genetics. Aca- Anat Rec 226:279–287
demic Press, Cambridge, MA, pp 1989–1990 7. Kuo M, Lan W, Lin S, Tsai K, Hahn L (1992)
5. Couble ML et al (2000) Odontoblast differenti- Collagen gene-expression in human dental-pulp
ation of human dental pulp cells in explant cell-cultures. Arch Oral Biol 37:945–952
cultures. Calcif Tissue Int 66:129–138 8. Gronthos S et al (2000) Postnatal human dental
6. Ishizeki K, Nawa T, Sugawara M (1990) Calcifi- pulp stem cells (DPSCs) in Vitro and in vivo.
cation capacity of dental papilla mesenchymal Proc Natl Acad Sci U S A 97(25):13625–13630.
Print
Chapter 4
Isolation of Dental Stem Cell-Enriched Populations

from Continuously Growing Mouse Incisors
Anamaria Balic
Abstract
Continuous growth of the rodent incisor is enabled by epithelial and mesenchymal stem cells (ESCs and
MSCs) which unceasingly replenish enamel and dentin, respectively, that wear by persistent animal gnaw-
ing. Lineage tracing studies have provided evidence that ESCs contribute to all epithelial lineages of the
tooth in vivo. Meanwhile, in the mouse incisor, MSCs continuously contribute to odontoblast lineage and
tooth growth. However, in vitro manipulation of ESCs has shown little progress, mainly due to lack of
appropriate protocol to successfully isolate, culture, expand, and differentiate ESCs in vitro without using
the co-culture system. In this chapter we describe the isolation of the Sox2-GFP+ cell population that is
highly enriched in ESCs. Isolated cells can be used for various types of analyses, including in vitro culture,
single cell-related analyses, etc. Furthermore, we describe ways to obtain populations enriched in the incisor
MSCs using FACS sorting of antibody-labeled cells. Easily accessible FACS sorting enables easy and
relatively fast isolation of the cells labeled by the fluorescent protein.
Key words Dental tissues, Stem cell-enriched population, Mouse incisors, Tooth regeneration
1 Introduction
In recent years, major effort has been made toward generating

dental tissues in vitro for the tooth regeneration purposes. The
major setback is the lack of appropriate cell source that would be
feasible for clinical use and our inability to induce odontogenic
potential in these cells. Continuous growth of the rodent incisor
is enabled by epithelial and mesenchymal stem cells (ESCs and
MSCs) which unceasingly replenish enamel and dentin, respec-
tively, that wear by persistent animal gnawing. Therefore, this
tooth is a perfect source of stem cells to study their fates and
differentiation in vivo, as well as to obtain them for further molec-
ular and cellular analyses in vitro.
ESCs reside in morphologically distinct cervical loops located
at the proximal end of the incisor and are identified as label-
retaining cells expressing Sox2, Bmi1, Oct4/3, Lgr5, and Yap
29
30 Anamaria Balic
[1–4]. Lineage tracing studies have provided evidence that ESCs

contribute to all epithelial lineages of the tooth in vivo
[5–7]. Extensive studies on mouse incisor have unraveled some of
the molecular regulatory networks which govern in vivo mainte-
nance and differentiation of ESCs into various epithelial cell
lineages. However, in vitro manipulation of ESCs has shown little
progress, mainly due to lack of appropriate protocol to successfully
isolate, culture, expand, and differentiate ESCs in vitro without
using the co-culture system.
In contrast, in vitro odontoblast differentiation and dentin
production have been analyzed extensively in the past decades
using dental pulp tissue obtained from various species and mainly
from teeth which seize to grow. These teeth are abundant in the
progenitor cells and contain very low number of MSCs that are
involved in reparative dentinogenesis [8]. Meanwhile, in the mouse
incisor, MSCs continuously contribute to odontoblast lineage and
tooth growth. MSCs are mainly recruited into dental pulp from the
neurovascular bundle, a network of blood vessels and mandibular
nerve that penetrates incisor dental pulp and brings various MSC
populations, including nerve-associated glial cells and pericytes
[9–11]. Some of the recruited cells reside in the mesenchyme
spanning the lingual and labial cervical loops as Thy1+ (CD90+)
and label-retaining cells [10].
In recent years, transgenic mice carrying GFP coding sequences
under the control of cell-specific regulatory elements have emerged
as a powerful tool for developmental and lineage studies of the
specific cell populations, including stem cells. Easily accessible
FACS sorting enables easy and relatively fast isolation of the cells
labeled by the fluorescent protein. In this chapter we describe the
isolation of the Sox2-GFP+ cell population that is highly enriched
in ESCs. Isolated cells can be used for various types of analyses,
including in vitro culture, single cell-related analyses, etc. Further-
more, we will describe ways to obtain populations enriched in the
incisor MSCs using FACS sorting of antibody-labeled cells.
2 Materials
2.1 Reagents 1. PBS.

2. Hank’s balanced salt solution (Sigma, cat. no. H 4034).
3. Collagenase P (Roche, CAT#11213857001).
4. Dispase (Sigma, D4693).
5. Pancreatin (Sigma, P-3292).
6. Trypsin (Difco, 215240).
7. NaCl.
8. KCl.
Isolation of Dental Stem Cell Enriched Populations from Mouse Incisors 31
9. NaH2PO4+H2O.
10. Glucose.
11. NaHCO3.
12. Sodium acetate.
13. Calcium acetate.
14. Fetal calf/bovine serum.
15. HEPES.
16. Propidium iodide.
17. DMEM.
2.2 Media and 1. Dissecting media: 1HBSS, 10 mM HEPES, pH 7.2.

Solutions 2. Dispase solution: prepare dispase stock solution (activity
50 U/mL) in dispase buffer (10 mM NaOAc, 5 mM CaOAc,
pH 7.4). Can be stored at +4 C for a month.
3. Pancreatin, 10 stock solution: 25 g/L pancreatin, 8.5 g/L
NaCl in Millipore water, pH 7.0; filter sterilize. Store in 1 ml
aliquot at 20 C.
4. Pancreatin-trypsin solution: 1 mL of pancreatin stock solution,
0.225 g of trypsin; dissolve in 6 mL of Tyrode solution. Store
in 1 mL aliquot at 20 C.
5. Recovery media: 10% fetal bovine (calf) serum in DMEM.
6. Collagenase solution: 0.5–3 U/mL collagenase P in PBS.
7. Sorting media: 1HBSS; 2% fetal bovine serum; 10 mM
HEPES, pH 7.2; and 1 μg/mL propidium iodide.
8. Tyrode solution (Ca2+, Mg2+-free): 8 g/L of NaCl, 0.2 g/L of
KCl, 0.005 g/L of NaH2PO4 + H2O, 1 g/L of glucose, and
1 g/L of NaHCO3. Filter sterilize and store at +4 C.
3 Methods
Sacrifice the mice following an established and approved animal

care protocols, and decapitate them. Separate the mandible from
the rest of the head, and place it on a Petri dish containing PBS.
Continue until all the mandibles are isolated. Split the mandible in
two by cutting through cartilage symphysis, and clean the soft
tissue as much as possible. Note that mandibles obtained from
P10 and older mice are mineralized and easy to clean, while those
obtained from younger pups are softer and more fragile (Fig. 1).
Separate the coronoid process from alveolar bone using twee-
zers. The bone breaks easily if the coronoid part in the close
proximity to the third molar is pushed toward the buccal side,
using tweezers. Most often the coronoid process of the mandible
32 Anamaria Balic
Fig. 1 Dissection of the neurovascular bundle and the apical end of the incisor from murine mandible.
Mandible of an 8-week-old mouse has been isolated and cleaned from the surrounding tissue (a). Breaking
the coronoid process exposed the neurovascular bundle (white arrow) and the remaining of mandibular nerve
(n. mandibularis) that is part of the neurovascular bundle (b). Neurovascular bundle is gently pulled away and
cut as indicated by yellow line (c). The apical end of the incisor is clearly visible (d), and labial cervical loop can
be observed (e)
will break as a plate exposing the entire proximal end of the incisor,
as well as the neurovascular bundle. The following text describes
how to separately isolate populations of interest. Just small mod-
ifications of the protocol are required to enable isolation of all of
them from the same mandible.
3.1 Isolation of the 1. Push the apical end of the incisor away from the bone, and
Incisor ESCs dissect the most proximal part containing cervical loop
(Fig. 1d, e).
2. There are at least three possible ways to obtain cervical loop:
one is mechanical separation and the other two involve enzy-
matic separation using different enzymes:
(a) Mechanical separation is best in cases where the time for
isolation is limited. If performed with great care, the
contamination with the adjacent mesenchyme is minimal.
Use fine tweezers to gently pull the cervical loop away
from the dental pulp mesenchyme. Once the cervical loop
is obtained, dissect out as much of the differentiating
epithelial tissue (preameloblasts and ameloblast layers,
stratum intermedium, etc.). Collect the cleaned cervical
loops in PBS.
(b) Enzymatic separation can be performed using dispase
enzymatic solution or pancreatin-trypsin. Always use
glass dishes.
When using dispase, prepare a working solution by
diluting the stock solution in the Opti-MEM to obtain the
final activity of 2–2.5 U/mL. Incubate the apical ends of
the incisor with dispase on room temperature for
20 min–1 h, with gentle rocking. Check every 20 min if
the epithelium is separating. Use fine tweezers or 28G
needle to separate the cervical loops. Transfer the cervical
loops to the new Petri dish with the clean PBS to stop the
enzyme activity.
Pancreatin-trypsin treatment takes 1–8 min at +37 C
to effectively separate the epithelium from the mesen-
chyme. Collect the separated cervical loops into a glass
dish containing enzymatic solution, and swirl occasionally
at +37 C. When you observe that cervical loops are
detaching from the mesenchyme, transfer them into
recovery media containing DMEM and 10% fetal bovine
(or calf) serum. The presence of serum stops the enzyme
activity and also allows the tissue to recover. Keep the
tissue in the recovery media for up to an hour, and then
use fine tweezers or 28G needles to separate the cervical
loops completely from the mesenchyme of the dental pulp
(see Note 1).
3. Place the isolated and cleaned cervical loops in collagenase
solution (activity 3 U/mL), and incubate at 37 C with gentle
rocking for 15–45 min (see Note 2).
4. Stop the enzymatic dispersion by addition of the fetal bovine
(or calf) serum. 5–10% fetal bovine serum is sufficient to
34 Anamaria Balic
inactivate the enzymes. Gently mix the samples and spin at

200–300 RCF for 10 min at +4 C.
5. Mechanically break loose tissue fragments using trituration
method to ensure optimal yield of cells.
6. Strain the cells through a 70 μm strainer, and count them using
dyes such as trypan blue to exclude dead cells.
7. Centrifuge the cells at 200–300 RCF for 10 min at +4 C, and
reconstitute the cell pellet in sorting media to reach a cell
density between 0.5 106 and 1 106 cells/mL to ensure
that cells do not adhere together until sorting is completed (see
Note 3).
8. Proceed with sorting according to the regulations dictated by
your FACS Core Facility, and collect the cells in media deter-
mined by the type of the study they will be used for.
3.2 Isolation of Cell After breaking the coronoid bone of the mandible and exposing the
Populations Enriched entire proximal end of the incisor (Fig. 1b), use tweezers to push
in MSCs from the the apical end of the incisor away from the bone. The neurovascular
Neurovascular Bundle bundle, which can be observed capping the apical end of the
incisor, is easily removed (Fig. 1b, c):
1. Collect the isolated bundles into collagenase solution
(0.5 U/mL), and incubate at 37 C with gentle rocking for
no more than 15 min.
2. Use gentle trituration method to further disperse the bundles
and to ensure optimal yield of cells.
3. Strain the tissue through a 40 μm strainer, and stop the enzy-
matic dispersion by addition of the fetal bovine (or calf) serum.
A 5–10% fetal bovine serum is sufficient to inactivate the
enzyme.
4. Gently mix the samples, and spin at 200–250 RCF for 10 min
at +4 C (see Note 4).
5. Reconstitute the cell pellet in 100–500 μL PBS at cell density of
0.5–1 106 cells/mL, and proceed with the antibody labeling
described in the following section.
3.3 Isolation of MSCs After breaking the coronoid process of the mandible and exposing
from the Incisor the entire proximal end of the incisor, use tweezers to push the
Dental Pulp apical end of the incisor away from the bone (Fig. 1), and dissect
the most proximal part containing cervical loop. Dental pulp mes-
enchyme spanning the loops is the location of the Thy1+, Gli1+,
and Axin2+, slow-cycling label-retaining MSCs.
1. Remove cervical loops as described in Subheading 3.1. The
best way is the mechanical separation by gentle pulling of the
labial cervical loop away from the dental pulp mesenchyme.
The lingual cervical loop is smaller and can be pinched off with
fine tweezers. Mechanical removal of cervical loops is some-
times followed by the removal of the small number of mesen-
chymal cells directly adjacent to the cervical loops. However,
these are most likely cells already committed to odontoblast
lineage and not the focus of this protocol.
2. Collect the cleaned apical mesenchyme in a 15 mL tube con-
taining collagenase P solution (1.5 U/mL) for enzymatic cell
dispersion. Pulp tissue dispersion is performed at 37 C with
gentle rocking for no more than 45 min. After the initial
15 min, the solution becomes cloudy due to active tissue
dissociation and extracellular matrix breakdown, indicative of
optimal enzymatic activity necessary for successful cell harvest.
3. Enzymatic dispersion is stopped by addition of the fetal bovine
(or calf) serum. A 5–10% fetal bovine serum is sufficient to
inactivate the enzymes. Gently mix the samples and spin at
200–300 RCF for 10 min at +4 C.
4. Mechanically break loose tissue fragments using trituration
method to ensure optimal yield of cells.
5. Strain the cells through a 70 μm strainer, and count them using
dyes such as trypan blue to exclude dead cells.
6. Centrifuge the cells at 200–300 RCF for 10 min at +4 C, and
reconstitute the cell pellet in 100–500 μl of PBS at cell density
of 0.5–1 106 cells/mL.
7. Proceed with antibody labeling (see Note 5).
The following protocol describes the procedure for the use of
conjugated antibodies.
8. To the tube with cells, add primary antibody in a correct
dilution that has been predetermined.
9. Gently mix and place the tube on ice for 45 min in dark.
10. Add fresh PBS and spin the cells at 300 RCF for 5 min.
11. Discard the PBS and repeat the wash at least once more.
12. Resuspend the cells in sorting media, and proceed with sorting
according to the regulations dictated by your FACS Core
Facility (see Note 6).
4 Notes
1. Pancreatin-trypsin enzymatic dispersion is timewise a long

procedure, but it ensures complete separation of the cervical
loops with no mesenchymal cell contamination.
2. If the cervical loops were separated mechanically, they will
require longer time to enzymatically disperse to single-cell
solution.
36 Anamaria Balic
3. For the proper FACS sorting, it is recommended to have a

negative control, as well as the control for each of the fluor-
ophores used. To ensure the optimal survival of the cells during
sorting, media with higher concentration of the fetal calf
(bovine) serum can be used. However, cell adherence can
increase and repeated filtering might be required.
4. Spinning the sample at low speed minimizes the amount of
dispersed nerve tissue that is a by-product of enzymatic disper-
sion and can interfere with the survival of the cells.
5. Most of the antibodies used for the FACS or flow cytometry
analyses are already conjugated. However, there are occasion-
ally those that require secondary antibody labeling.
6. Strategy for FACS sorting: Most often FACS sorting is per-
formed by appropriately trained personnel who will ensure
that the cell sorter is optimized and that it is sorting accurately.
Select the appropriate nozzle dependent on the type of cells
that are sorted. To determine an accurate fluorescence gating
scheme, control samples are necessary to distinguish between
the true and false-positive and false-negative cells. These con-
trols include:
(a) The full negative sample obtained from the same tissue
from the non-transgenic animal using the same protocol
and placed in the sorting media with no propidium iodide
(PI)
(b) Control for fluorophore used: sample of the same tissue
obtained from non-transgenic animal and placed in the
sorting media with no PI
(c) Control for PI: sample of the same tissue obtained from
non-transgenic animal using the same protocol and placed
in the sorting media with PI
Following the FACS sorting, a reanalysis of the sorted
samples reassures the purity of the sorted populations that
should not be below 97.5%.
References
1. Balic A, Thesleff I (2015) Tissue interactions developing mouse incisor. Gene Expr Patterns:
regulating tooth development and renewal. GEP 11(3–4):163–170
Curr Top Dev Biol 115:157–186 4. Suomalainen M, Thesleff I (2010) Patterns of
2. Harada H, Kettunen P, Jung HS, Mustonen T, wnt pathway activity in the mouse incisor indi-
Wang YA, Thesleff I (1999) Localization of cate absence of wnt/beta-catenin signaling in
putative stem cells in dental epithelium and the epithelial stem cells. Dev Dyn 239
their association with notch and fgf signaling. (1):364–372
J Cell Biol 147(1):105–120 5. Biehs B, Hu JK, Strauli NB, Sangiorgi E,
3. Li L, Kwon HJ, Harada H, Ohshima H, Cho Jung H, Heber RP, Ho S, Goodwin AF,
SW, Jung HS (2011) Expression patterns of Dasen JS, Capecchi MR et al (2013) Bmi1
abcg2, bmi-1, oct-3/4, and yap in the represses ink4a/arf and hox genes to regulate
stem cells in the rodent incisor. Nat Cell Biol erupted and unerupted murine molars. Bone
15(7):846–852 46(6):1639–1651
6. Juuri E, Saito K, Ahtiainen L, Seidel K, 9. Feng J, Mantesso A, De Bari C, Nishiyama A,
Tummers M, Hochedlinger K, Klein OD, Sharpe PT (2011) Dual origin of mesenchymal
Thesleff I, Michon F (2012) Sox2+ stem cells stem cells contributing to organ growth and
contribute to all epithelial lineages of the tooth repair. Proc Natl Acad Sci U S A 108
via sfrp5+ progenitors. Dev Cell 23 (16):6503–6508
(2):317–328 10. Kaukua N, Shahidi MK, Konstantinidou C,
7. Seidel K, Ahn CP, Lyons D, Nee A, Ting K, Dyachuk V, Kaucka M, Furlan A, An Z,
Brownell I, Cao T, Carano RA, Curran T, Wang L, Hultman I, Ahrlund-Richter L et al
Schober M et al (2010) Hedgehog signaling (2014) Glial origin of mesenchymal stem cells
regulates the generation of ameloblast progeni- in a tooth model system. Nature 513
tors in the continuously growing mouse inci- (7519):551–554
sor. Development 137(22):3753–3761 11. Zhao H, Feng J, Seidel K, Shi S, Klein O,
8. Balic A, Aguila HL, Caimano MJ, Francone Sharpe P, Chai Y (2014) Secretion of shh by a
VP, Mina M (2010) Characterization of stem neurovascular bundle niche supports mesen-
and progenitor cells in the dental pulp of chymal stem cell homeostasis in the adult
mouse incisor. Cell Stem Cell 14(2):160–173
Chapter 5
Application of Cell Lineage Tracing Combined with

Immunofluorescence in the Study of Dentinogenesis
Yan Jing, Chaoyuan Li, and Jian Q. Feng
Abstract
The cell lineage tracing system has been used predominantly in developmental biology studies. The Cre
recombinase allows for the activation of the reporter in a specific cell line and all progeny. In this protocol,
we will introduce how the cell lineage tracing technique can be performed in the investigation of dentino-
genesis by using Gli1-CreERT2; R26RTomato compound mice. Moreover, we combined cell lineage tracing
in conjunction with immunofluorescence—to further define cell fate by analyzing the expression of specific
cell markers for odontoblasts. This combination not only broadens the application of cell lineage tracing but
also simplifies the generation of compound mice. More importantly, the number, location, and differentia-
tion status of parent cell progeny can be displayed simultaneously, providing more information than cell
lineage tracing or immunofluorescence alone. In conclusion, the co-application of cell lineage tracing
technique and immunofluorescence is a powerful tool for investigating cell biology in the field of dentino-
genesis and tooth development.
Key words Cell lineage tracing, Immunofluorescence, Odontoblast, Dentin tubule, Dentinogenesis,
Gli1
1 Introduction
It is commonly accepted that the reciprocal interaction between the

epithelium and mesenchyme is essential for odontoblast differenti-
ation and odontoblast process formation [1, 2]. However, it is hard
to explain how postnatal dentin masses are explosively expanded in
the absence of the reciprocal interaction between epithelial and
mesenchymal cells due to newly formed enamel and dentin layers.
In order to answer the question, it is necessary to dynamically track
the cell fate and reveal the source of newly formed odontoblasts and
dentin tubules during the postnatal growth of dentin.
The cell lineage tracing system, which has been used predomi-
nantly in developmental biology studies, provides a more rigorous
way to study cell fate. The use of Cre recombinase allows for the
activation of the reporter in a specific cell line and all progeny.
39
40 Yan Jing et al.
Briefly speaking, the Cre recombinase enzyme, which is only

expressed in a specific type of cell, stimulates the expression of the
reporter gene. In this way, this type of cell and their descendants are
permanently labeled [3]. In some cases, the investigator can choose
a favorable time point to activate Cre by using a drug, such as
tamoxifen, when Cre is fused to a modified form of the estrogen
receptor (CreERT2) (see Note 1) [4]. Fluorescent reporters have
become the standard in lineage tracing experiments because they
dramatically reduce complexity and improve the accuracy and effi-
ciency of cell fate tracing [4, 5]. tdTomato is the best choice among
fluorescent reporters since it has the brightest fluorescent protein
and strongest epifluorescence, making it easily visualized [3].
In this protocol, we will introduce how the cell lineage tracing
technique can be used in the study of dentinogenesis. Moreover, we
will combine lineage tracing with immunofluorescence, which can
further define cell fate by analyzing the expression of specific cell
markers for odontoblasts. This combination broadens the applica-
tion of cell lineage tracing by simplifying the generation of com-
pound mice and providing more valuable information, such as the
number, location, and differentiation status of parent cell
progeny [6].
2 Materials
2.1 Sample 1. 10 mg/mL tamoxifen (see Note 2).

Preparation 2. 28G 1/2 syringe.
3. 4% paraformaldehyde (PFA) solution in phosphate-buffered
saline (PBS), pH 7.4 (see Note 3).
4. 10% EDTA, pH 7.4.
5. 15–30% sucrose solution in PBS.
6. Dissection scissors; #3 and #5 forceps.
7. 70% ethanol.
8. 50 mL polypropylene centrifuge tube.
9. Frozen cryosection machine.
10. Glass slides.
2.2 Slide Preparation 1. Immunofluorescence antigen retrieval: hyaluronidase powder

for Confocal Image in PBS at a concentration of 2 mg/mL, pH 5.0.
2. Blocking solution: 3% BSA (bovine serum albumin) and 20%
goat serum in PBS.
3. Primary antibody solution: Nestin antibody (Millipore, 1:100)
and 2% goat serum in PBS.
Cell Lineage Tracing in Dentinogenesis 41
4. Secondary antibody solution: secondary antibody with a ratio

of 1:500 and 2% goat serum in PBS.
5. Hydrophobic barrier pen.
6. Distilled water.
7. PBS.
8. 50 mL glass jars.
9. 1.5 mL Eppendorf tube.
10. Pipette.
11. DAPI.
2.3 Fluorescent Confocal Microscope and Imaging Center.

Imaging
3 Methods
3.1 Sample 1. Cross Gli1-CreERT2 mice [7] with R26RTomato (B6; 129S6-Gt
Preparation for Cell (ROSA)26Sortm9(CAG-tdTomato)Hze/J) mice to obtain Gli1-
Lineage Tracing CreERT2; R26RTomato mice (see Note 4).
2. Inject tamoxifen for Gli1-CreERT2; R26RTomato mice at a favor-
able time point (see Note 5). First, remove the mouse from the
cage. Then, use the left thumb and index finger to grab the skin
on the back of the mouse, and turn it over, exposing the
abdomen. Use the right hand to hold the syringe. The optimal
entry point for injection is on the left or right side of the
hypogastrium, avoiding the liver and bladder. Keep the syringe
parallel to the hind legs of the mouse and inject intraperitone-
ally. The dosage for injection is 75 mg/kg (see Note 6).
3. Choose a favorable time point to harvest mice based on the
study aim (see Note 7). In this protocol, we injected the
tamoxifen at postnatal day 3 (P3).
4. On the scheduled harvest time, euthanatize the mice by CO2
(see Note 8). In this protocol, we harvested the mice at post-
natal days 4 (P4) and 17 (P17), respectively.
5. Peel off the mouse’s skin, and put the whole body into a 50 mL
polypropylene centrifuge tube that contains 40 mL 4% PFA to
fix overnight at 4 C.
6. Use dissection scissors and #3 and #5 forceps to carefully
remove the mandible and get rid of the muscles on the surface.
7. Put the mandible into 10% EDTA to decalcify at 4 C in a
50 mL polypropylene centrifuge tube (see Note 9).
8. Use 50 mL 15% sucrose to dehydrate the mandible overnight
at 4 C in a 50 mL polypropylene centrifuge tube.
42 Yan Jing et al.
9. Use 50 mL 30% sucrose to dehydrate the mandible overnight

at 4 C in a 50 mL polypropylene centrifuge tube.
10. Embed the mandible with OCT on the cutting plate in the
cryosection machine. First, put the mandible on the mounting
mold in the favorable direction. Next, submerge the tissue in
OCT, and leave it in the frozen cryosection machine till the
OCT freezes. Afterward, mount the OCT block on the cutting
plate. Wait for approximately 15 min before cutting to ensure
the OCT is completely frozen (see Note 10).
11. Cut the teeth into 5–10 μm sections (see Note 11). Collect the
sections on slides and store at 20 C.
12. Incubate the slide in a 37 C chamber for 10 min to remove the
water before starting to stain.
13. Use 50 mL glass jar to wash the slide twice with distilled water
to get rid of the OCT, 5 min per time.
14. Wipe off the water around the section. Use a hydrophobic
barrier pen to circle the section, and drop DAPI or
non-fluorescing antifade mountant into the circle. Carefully
lay down the cover slip (see Note 12).
3.2 Sample 1. Incubate the slides for 10 min in a 37 C chamber to remove

Preparation to the water before staining.
Combine the Cell 2. Use 50 mL glass jar to wash the slides twice with distilled water
Lineage Tracing with to get rid of the OCT, 5 min per time.
Immunofluorescence 3. Use the hydrophobic barrier pen to circle each section on the
slide. From this step, add all of the prepared solution within the
circle to completely cover sections (see Note 12).
4. Treat the sections with hyaluronidase in a humid chamber at
37 C for 30–60 min (see Note 13). The volume of the solution
from steps 4 to 7 depends on the size of the section. 25 μL of
solution is necessary for mice tooth. Wash with PBS three
times, 3 min per time (see Note 14).
5. Prepare and apply the blocking solution to each section, and
incubate in a humid chamber for 1 h at room temperature.
6. Incubate the sections with primary antibody solution at 4 C
overnight. Wash with PBS three times, 3 min per time. In
this protocol, we will use anti-mouse Nestin antibody (see
Note 15).
7. Incubate the sections with secondary antibody solution for 2 h
at room temperature. Wash with PBS three times, 3 min
per time.
8. Wipe off the water around the section, and drop the DAPI into
the circle to cover the section on the slide. Carefully lay down
the cover slip.
Fig. 1 The co-application of Nestin immunofluorescence with lineage tracing background in the first molar of
4-day Gli1-CreERT2; R26RTomato compound mice (tamoxifen was injected at postnatal day 3). (a) The tomato
signal reflected only a few of Gli1+ cells represented by red color in pulp (arrows). (b) The Nestin immunofluo-
rescence signal reflected its expression in odontoblast processes in the predentin layer. (c) The co-localization
of Nestin expression with tomato signal showed that most of the dentin tubules were in green color, which
further confirmed that very few of odontoblasts differentiated from Gli1+ cells and formed their processes
(arrows) (Tm: tamoxifen)
44 Yan Jing et al.
3.3 Confocal 1. Capture fluorescent cell images using a confocal microscope.

Microscopy Take multiple stacked images at 200 Hz (dimensions of
1024 1024), using 10, 20, and 63 lenses [8].
2. At P4, the cell lineage tracing result showed that the majority of
odontoblasts were in blue color, indicating that those cells were
differentiated from the pulp progenitors before the tamoxifen
injection. Only a few of Gli1+ cells in red color appeared the
odontoblast layers in the pulp, which were the ones tagged by
tamoxifen injection at P3 (Fig. 1a, arrows). However, many
more red odontoblasts and dentin tubules were formed and
represented in red color at P17 (14 days after tamoxifen injec-
tion) (Fig. 2a). The increase of tomato signal not only reflects
the newly differentiated odontoblasts and freshly formed den-
tin tubules but also identifies their origin—the Gli1+ progeni-
tors in dental pulp. However, none of these essential evidences
can be revealed by Nestin immunofluorescence only, which
merely displayed similar expression level and location of Nestin
at P4 and P17 (Figs. 1b and 2b). Thus, compared with the
traditional assays, such as immunofluorescence, the cell lineage
tracing technique provides an intuitive understanding of the
cell source for the postnatal dentinogenesis.
3. More importantly, if we combine the cell lineage tracing tech-
nique with Nestin immunofluorescence, it is further confirmed
that few dentin tubules were formed by Gli1+ cell-derived
odontoblasts at P4 since most odontoblast processes were
pure green (Fig. 1c), but a large number of new dentin tubules
were established by Gli1+ cell-derived odontoblasts at P17
(Fig. 2c). In other words, the co-application of the cell lineage
tracing and the immunofluorescence is better than using either
of them alone, because it can reflect the number, location, and
differentiation status of parent cell progeny simultaneously,
providing more information and simplifying the generation of
compound mice.
4 Notes
1. Compared with non-inducible Cre system, the activation of

inducible Cre can be restricted spatially and temporally [9].
2. Dissolve the tamoxifen powder in 1/10 volume of ethanol and
9/10 volume of corn oil at a concentration of 10 mg/mL.
Stock at 20 C. It is better to warm the 100% ethanol at 55 C
for 30 min before dissolving. If tamoxifen is not dissolved well,
transfer all the ethanol and undissolved tamoxifen into a bea-
ker, add moderate coin oil, and warm up the beaker at 55 C for
30 min.
Fig. 2 The co-application of Nestin immunofluorescence with lineage tracing background in the first molar of
17-day Gli1-CreERT2; R26RTomato compound mice (tamoxifen was injected at postnatal day 3). (a) The tomato
signal showed many more Gli1+ cell-derived odontoblasts and newly formed dentin tubules than P4. (b) The
Nestin immunofluorescence signal reflected its expression in odontoblast processes in the predentin layer. (c)
The co-localization of Nestin expression with tomato signal displayed that a majority of odontoblasts and
dentin tubules were labeled by both immunofluorescent (green) and tomato (red) signal 14 days after
tamoxifen injection (Tm: tamoxifen)
46 Yan Jing et al.
3. Since PFA has toxicity, handle it in the hood with gloves and
facemask.
4. In this protocol, we use Gli1-CreERT2 as an example to show
how cell lineage tracing and its co-application with immuno-
fluorescence can be used in the study of dentinogenesis. The
investigators can choose other types of Cre with different
activation ways (inducible or non-inducible) due to the study
goals, such as Osx-Cre, 2.3Col1a1-Cre, and 3.2Col1a1-CreERT2
[10, 11].
5. For inducible Cre system, the investigator can select the injec-
tion time according to the study aims and the expression time
of the tagged gene. For example, Gli1 is a transcriptional factor
mainly expressed in early progenitors. Thus, a large number of
odontoblasts will be labeled in red color if activating Gli1-
CreERT2 in the tooth development stage for a period of time
(Fig. 2). However, fewer will be labeled if Gli1-CreERT2 is
activated after the tooth is well developed.
6. The working range for tamoxifen is 75–300 mg/kg. It has
been reported that the dose of tamoxifen may change the
efficiency of Cre activation and the number of labeled cells.
Low doses will label the population of interest at clonal density
[12]; high doses may label the entire progenitor pool
[3, 13]. Thus, the dose must be chosen depending on the
purpose of the experiment. In addition, tamoxifen has poten-
tial toxicity, especially in high doses [14, 15]. It is better to
decrease the dose to avoid late-term abortions when injecting
during pregnancy (100 μL per pregnant mouse) [16].
7. If cell proliferation assays are required, EdU or BrdU can be
injected before harvest: for EdU, one injection at 2 h before
sacrifice and for BrdU, two injections at 24 h and 2 h separately
before sacrifice.
8. The use of CO2 is accepted for euthanasia in mice by the
American Veterinary Medical Association (AVMA). Based on
the AVMA recommendations, compressed CO2 gas in cylin-
ders will be used, and the optimal flow rate used will displace at
least 20% of the chamber volume per minute.
9. The duration of decalcification is variable due to the size and
age of the tooth. The older the mice, the longer it takes. There
are three ways to accelerate decalcification: (a) cut off the
anterior part of the incisor and the posterior part of the mandi-
ble (distal from the third molar), to speed up the penetration of
EDTA; (b) use enough EDTA and change it regularly; and (c)
prepare EDTA in a higher concentration, such as 10–17%.
10. Traditionally, there are two common ways to cut the tooth:
along with sagittal plane and coronal plane. The investigator
can choose one based on the study interest. Cutting mice tooth
requires accurate embedding position. In order to keep the

sample in the ideal orientation when using OCT to embed, we
suggest to put the mold on dry ice. In this way, the sample can
be frozen quickly in OCT, so that its position will not change
during embedding.
11. The investigator can adjust the thickness of each section due to
experiment requirements. For example, thinner section
(4–5 μm) is favorable for histological staining, such as H&E.
12. For DAPI staining, we recommend to use Life Technologies
(P36931), which is easy to handle. When doing incubation
during immunofluorescence, or mounting slides with DAPI,
we recommend to use hydrophobic barrier pen to circle the
sample first, which can save the reagent and guarantee the
incubation efficiency.
13. The incubation time is variable. Incubate 30 min for samples
less than 3 weeks old and 40–60 min for those older than
3 weeks old.
14. When doing the immunofluorescence for the markers
expressed in the nucleus, such as transcriptional factors, we
recommend to use PBST (PBS that contains 0.1% Tween 20)
to wash the sample before blocking.
15. Nestin is an intermediate filament protein and a marker of bone
and odontoblast cells. Thus, in this protocol, we used Nestin as
an example to show how to combine the cell lineage tracing
with immunofluorescence in dentinogenesis study. In order to
avoid false-positive results during immunofluorescence, we
suggest to use IgG as the primary antibody in the control
staining and perform it with experimental staining
simultaneously.
References
1. Thesleff I (2003) Epithelial-mesenchymal sig- 6. Jing Y, Hinton RJ, Chan KS et al (2016)
nalling regulating tooth morphogenesis. J Cell Co-localization of cell lineage markers and the
Sci 116:1647–1648 tomato signal. J Vis Exp (118)
2. Li J, Parada C, Chai Y (2017) Cellular and 7. Zhao H, Feng J, Ho TV et al (2015) The
molecular mechanisms of tooth root develop- suture provides a niche for mesenchymal stem
ment. Development 144:374–384 cells of craniofacial bones. Nat Cell Biol
3. Kretzschmar K, Watt FM (2012) Lineage 17:386–396
tracing. Cell 148:33–45 8. Ren Y, Lin S, Jing Y et al (2014) A novel way to
4. Humphreys BD, DiRocco DP (2014) Lineage- statistically analyze morphologic changes in
tracing methods and the kidney. Kidney Int Dmp1-null osteocytes. Connect Tissue Res
86:481–488 55(Suppl 1):129–133
5. Romagnani P, Rinkevich Y, Dekel B (2015) 9. Pest MA, Beier F (2014) Developmental biol-
The use of lineage tracing to study kidney ogy: Is there such a thing as a cartilage-specific
injury and regeneration. Nat Rev Nephrol knockout mouse. Nat Rev Rheumatol
11:420–431 10:702–704
48 Yan Jing et al.
10. Zhang H, Jiang Y, Qin C et al (2015) Essential 14. Huh WJ, Khurana SS, Geahlen JH et al (2012)
role of osterix for tooth root but not crown Tamoxifen induces rapid, reversible atrophy,
dentin formation. J Bone Miner Res and metaplasia in mouse stomach. Gastroenter-
30:742–746 ology 142(21–24):e27
11. Canalis E, Parker K, Feng JQ et al (2013) 15. Lee MH, Kim JW, Kim JH et al (2010) Gene
Osteoblast lineage-specific effects of notch acti- expression profiling of murine hepatic steatosis
vation in the skeleton. Endocrinology induced by tamoxifen. Toxicol Lett
154:623–634 199:416–424
12. Rios AC, Fu NY, Lindeman GJ et al (2014) In 16. Nakamura E, Nguyen MT, Mackem S (2006)
situ identification of bipotent stem cells in the Kinetics of tamoxifen-regulated Cre activity in
mammary gland. Nature 506:322–327 mice using a cartilage-specific CreER(T) to
13. Blanpain C, Simons BD (2013) Unravelling assay temporal activity windows along the
stem cell dynamics by lineage tracing. Nat Rev proximodistal limb skeleton. Dev Dyn
Mol Cell Biol 14:489–502 235:2603–2612
Chapter 6
Tissue Recombination and Kidney Capsule Transplantation

Assays for the Study of Epithelial-Mesenchymal
Interactions
Lucia Jimenez-Rojo and Thimios A. Mitsiadis
Abstract
Tissue interactions are crucial during the development of organs. Among the most studied tissue interac-
tions are those that take place between the epithelial cells and the underlying mesenchymal cells, known as
epithelial-mesenchymal interactions. Tissue recombination assay is a valuable model to study the mechan-
isms involved in the regulation of such interactions. Here, we describe how to dissociate and recombine the
epithelial and mesenchymal components of the embryonic tooth. In addition, we explain how to transplant
the recombined tissues under the kidney capsule of immunocompromised mice in order to allow their
further development into a mature tooth.
Key words Tissue recombination, Tooth, Embryo, Kidney capsule transplantation, Epithelial-mesen-
chymal interactions, Organogenesis
1 Introduction
Ectodermal appendages (tooth, salivary glands, hairs, mammary

glands, etc.) develop through similar cellular mechanisms that
involve an intimate and controlled cross talk between the epithelial
and mesenchymal tissues [1]. When these interactions are disrupted
due, for example, to the deregulation of important signaling mole-
cules, the development of the organ is altered. In developmental
biology, tissue recombination and transplantation assays have been
proved as useful tools for the study of tissue interactions [2–4].
Both in vitro tissue recombination and in vivo transplantation
assays are useful when combined but also separately. Variations of
tissue recombination assay can be used to study different aspects of
the regulation of tissue interactions. For instance, heterotypic
(combination of tissues from different organs) [5, 6] or hetero-
chronic (combination of tissues from different developmental time
points) [7, 8] recombinations can be performed in order to study
49
50 Lucia Jimenez-Rojo and Thimios A. Mitsiadis
either the timing or the organ specificity of the interactions,

respectively.
On the other hand, transplantation of tissues or organs into a
well-vascularized organ (such as the kidney) provides a suitable
trophic support that allows their terminal maturation. For instance,
the whole embryonic teeth can be transplanted under the kidney
capsule and kept there until they reach maturation stage. This can
be especially useful when the mutant mice lacking the molecule
under investigation die embryonically before the teeth are
completely developed [9].
Here we describe a detailed protocol for the dissection of
embryonic teeth followed by dissociation/recombination of their
epithelial and the mesenchymal components. Then, we describe
how these recombinant dental tissues can be implanted under the
kidney capsule of mice in order to allow their development into an
entire mature tooth.
2 Materials
2.1 Dissection of 1. PBS 1.

Embryonic Teeth and 2. Plastic (polystyrene) petri dishes (90 mm and 35 mm
Tissue Dissociation/ diameters).
Recombination
3. Glass petri dishes (90 mm diameter).
4. Dissection instruments (scissors, tweezers, 25G sterile hypo-
dermic needles).
5. Hank’s Balanced Salt Solution (HBSS).
2.2 Dissociation/ 1. Digestion buffer: Dispase (2 mg/mL) and DNase (20 U/mL)
Reassociation of the solution in HBSS.
Epithelium and 2. PBS/10% fetal bovine serum (FBS): Mix 10 mL of FBS with
Mesenchyme 90 mL of PBS 1, and filter the solution using a 0.22 μm pore
size filter.
3. Semisolid medium: Add 1.8 mL of medium (DMEM/F12,
20% FBS, 1% penicillin/streptomycin, 1.8 mg/mL ascorbic
acid) to a plastic petri dish (35 mm), and place it onto a hot
plate (at around 55 C). Then prepare agarose at 5% by mixing
0.5 gr of agarose with 10 mL of distilled water in a previously
sterilized 50 mL Erlenmeyer flask. Heat the 5% agarose in the
microwave until the solution is well dissolved (see Note 1).
Then add 200 μL of 5% agarose to the pre-warmed medium
(see Note 2).
2.3 Transplantation 1. Buprenorphine: For subcutaneous injection, a stock solution of

Under the Kidney buprenorphine at 0.06 mg/mL will be prepared. For the over-
Capsule of Mice night pain management, 0.3 mg will be added to 160 mL of
drinking water.
Tissue Recombination and Kidney Capsule Transplantation Assays 51
2. Anesthesia machine with isoflurane.

3. Vitamin A ointment.
4. Warming pad.
5. Surgical material (tweezers, scissors, absorbable polyglycolic
acid suture, autoclip system, 9 mm metal clips).
6. Parafilm.
3 Methods
3.1 Dissection of 1. Sacrifice the pregnant mouse at embryonic day (E) 13.5.
Embryonic Mouse 2. Dissect out the uterus, put it in a plastic petri dish with cold
Molars from the Lower PBS, and place the petri dish on ice.
Jaw (Fig. 1)
3. Using scissors cut the uterus between the implantation sites to
separate the embryos.
4. Transfer one embryo into a petri dish with cold PBS.
5. Using forceps remove the uterus and the decidua.
6. Cut the umbilical cord to isolate the embryo from the yolk sac.
7. Transfer the embryo to a glass petri dish with cold PBS.
8. Using needles separate the head from the body.
9. Separate the upper and lower jaws.
10. Remove the tongue and separate the mandible in two hemi-
mandibles.
11. Dissect the molar, and place it in a petri dish (35 mm) with cold
HBSS (on ice) (see Note 3).
Fig. 1 Location of molars in E13.5 mouse embryos. After cutting off the head from the body (a), the lower and
upper mandibles are separated (b), and molars are visible in the lower mandible (c). Abbreviations: m molars,
t tongue
3.2 Dissociation/ 1. Add 100 μL of digestion buffer in a glass petri dish, and transfer
Recombination of the molars inside by using forceps (note: don’t press the
Epithelium and molars; they should be retained in between the tips of the
Mesenchyme (Fig. 2) forceps by capillarity).
2. Incubate for 20 min at room temperature (or 1 h at 4 C) (see
Note 4).
3. Transfer the molars to PBS/10% FBS to stop the enzymatic
activity.
4. Using dissection needles separate mechanically the epithelium
from the mesenchyme.
5. Place the mesenchyme on top of a plastic petri dish (35 mm)
with semisolid medium (see Note 5).
6. Put the epithelium on top of the mesenchyme (see Note 6).
7. Incubate the recombinants overnight at 37 C to let the tissues
to adhere.
Fig. 2 Dissociation and reassociation of the dental epithelium and mesenchyme

from E13.5 tooth germs. (a) Schematic representation of the dental epithelium
and the underlying mesenchyme in an E13.5 molar. (b) Dissociation of the
epithelium and mesenchyme after enzymatic and mechanical treatment. (c)
Dissociated mesenchyme is placed on top of a plastic petri dish with
semisolid medium, and then the epithelium is placed on top of the
mesenchyme. Abbreviations: ep epithelium, mes mesenchyme
Fig. 3 Transplantation of recombined dental mesenchyme and epithelium under

the kidney capsule of a mouse. (a) Tissue recombinants before transplantation
where the dental epithelium (dotted circle) can be observed on top of the dental
mesenchyme. (b) After exposing the kidney and making a small pocket under the
capsule, recombined tissues can be implanted. Abbreviations: ep epithelium,
mes mesenchyme, k kidney, p parafilm
3.3 Transplantation 1. Inject buprenorphine (0.1 mg/kg bodyweight) subcutane-

Under the Kidney ously to the recipient mice 1 hour before the surgery.
Capsule of Mice 2. Once the recipient mouse is anesthetized with the inhalation
(Fig. 3) anesthesia (isoflurane) (see Note 7), vitamin A ointment will be
applied to the eyes to prevent them from drying. Then, the
mouse will be put on one side, and an incision will be made that
allows the exposure of one kidney. A small incision will be made
to the kidney capsule using forceps. The capsule will be lifted
(see Note 8), and the recombinant to be grafted will be put
under the capsule (see Note 9) in such a way that the capsule
keeps the graft in place.
3. The animals will be then sutured with absorbable polyglycolic
acid suture, and the skin incision will be closed with a metal
clip. The mouse will be put onto a warming pad and observed
until it reaches consciousness.
4. Buprenorphine treatment will be applied the day of the surgery
and until the third day after the surgery. Posture, stiches, and
wound healing should be controlled during the days that fol-
low the surgery. If signs of distress become apparent (poor
appearance, inflammation at site of graft), the animal has to
be sacrificed prior to reaching the experiment end point.
4 Notes
1. Control that the solution doesn’t overflow during the heating.

2. When adding the dense agarose solution, cutting pipette tip is
recommended.
3. To transfer the molars to a plastic petri dish with HBSS, either a
yellow pipette tip or tweezers can be used. When using twee-
zers, it is important not to press the teeth; they should be held
with the tip of the tweezers with a bit of liquid (by capillarity).
4. Make sure that during the incubation time, the molars don’t
dry out. In dry environments, the digestion buffer may evapo-
rate; if so, replenish the evaporated liquid.
5. Use a yellow tip to transfer the mesenchyme. Once it is on top
of the semisolid medium, remove the excess of liquid carefully.
6. Use a yellow tip to transfer the epithelium. Since the transfer of
the epithelium is done with medium, it is possible that the
epithelium moves from the top of the mesenchyme and stays
in the semisolid plate. If this happens, take it together with
some liquid using tweezers, and place it back on top of the
mesenchyme. The tissues have to be always kept in a liquid
environment, since once they dry tissues become very sticky
and thus difficult to manipulate. After the epithelium is placed
on top of the mesenchyme, the excess of liquid around the
tissues can be removed.
7. Isoflurane will be used at 4–5% for induction and 1–3% for
maintenance.
8. A piece of parafilm can be used to stabilize the kidney (Fig. 3b).
9. Tissue recombinants can be grafted with a small piece of the
surrounding semisolid medium if that facilitates the insertion
under the kidney capsule.
References
1. Jimenez-Rojo L, Granchi Z, Graf D, Mitsiadis 4. Mitsiadis TA, Cheraud Y, Sharpe P, Fontaine-

TA (2012) Stem cell fate determination during Perus J (2003) Development of teeth in chick
development and regeneration of ectodermal embryos after mouse neural crest transplanta-
organs. Front Physiol 3:107. https://doi.org/ tions. Proc Natl Acad Sci U S A 100
10.3389/fphys.2012.00107 (11):6541–6545. https://doi.org/10.1073/
2. Kollar EJ, Baird GR (1969) The influence of the pnas.1137104100
dental papilla on the development of tooth shape 5. Sakakura T, Nishizuka Y, Dawe CJ (1976)
in embryonic mouse tooth germs. J Embryol Mesenchyme-dependent morphogenesis and
Exp Morphol 21(1):131–148 epithelium-specific cytodifferentiation in mouse
3. Kratochwil K (1969) Organ specificity in mesen- mammary gland. Science 194
chymal induction demonstrated in the embry- (4272):1439–1441
onic development of the mammary gland of the 6. Dhouailly D, Rogers GE, Sengel P (1978) The
mouse. Dev Biol 20(1):46–71 specification of feather and scale protein
synthesis in epidermal-dermal recombinations. induced by tissue recombination of embryonic

Dev Biol 65(1):58–68 pituitary epithelium and embryonic subman-
7. Mitsiadis TA, Graf D, Luder H, Gridley T, Blu- dibular gland mesenchyme in mice. Dev Biol
teau G (2010) BMPs and FGFs target Notch 110(2):382–391
signalling via jagged 2 to regulate tooth mor- 9. Caton J, Luder HU, Zoupa M, Bradman M,
phogenesis and cytodifferentiation. Develop- Bluteau G, Tucker AS, Klein O, Mitsiadis TA
ment 137(18):3025–3035. https://doi.org/ (2009) Enamel-free teeth: Tbx1 deletion affects
10.1242/dev.049528 amelogenesis in rodent incisors. Dev Biol 328
8. Kusakabe M, Sakakura T, Sano M, Nishizuka Y (2):493–505. https://doi.org/10.1016/j.
(1985) A pituitary-salivary mixed gland ydbio.2009.02.014
Part II
Protocols for Dental Stem Cells and Tissue Regeneration

Chapter 7
Dental Mesenchymal Stem Cells: Dental Pulp Stem Cells,

Periodontal Ligament Stem Cells, Apical Papilla Stem Cells,
and Primary Teeth Stem Cells—Isolation, Characterization,
and Expansion for Tissue Engineering
Mey Al-Habib and George T. -J. Huang
Abstract
Dental stem cells (DSCs) have been shown to possess great potential for multiple biomedical applications,
especially for dental tissue regeneration. They are a special type of subpopulation of mesenchymal stem/
stromal cells (MSCs) and present subtle differences from other types of MSCs. Therefore, it requires a
specialized expertise to isolate, culture, and characterize these cells in vitro and in vivo. The purpose of this
chapter is to share our experience in studying these cells. We will describe in detail laboratory protocols
outlining how the cells are isolated, cultured, expanded, and characterized using various in vitro cellular and
biochemical analyses, as well as an in vivo study model using immunocompromised mice to observe tissue
regeneration after transplantation of these DSCs.
Keywords Dental stem cells, Dental pulp, Periodontal ligament, Apical papilla, Primary teeth, Dental
tissues, Mesenchymal stem cells, In vitro and in vivo models, Tissue engineering and regeneration
1 Introduction
The discovery of dental stem cells (DSCs) in the 2000s has pro-
moted the research interest in not only the characterization of these
stem cells but also their applications in regenerative medicine.
DSCs are a novel population of mesenchymal stem/progenitor
cells that are highly proliferative and have the potential for self-
renewal and multi-lineage differentiation [1, 2]. Along with the
advancement of tissue engineering research, these DSCs have been
tested for regeneration of various dental and oral tissues such as the
pulp, dentin, cementum, periodontal ligament, and bone
[1, 3–5]. Additionally, DSCs have been found to be highly angio-
genic and neurogenic [6–8], and some reports have shown their
capacity of chondrogenesis albeit weak [9, 10]. Given that oral
tissues are easily accessible by dentists, DSCs can be feasibly
59
60 Mey Al-Habib and George T. -J. Huang
obtained from those tissues and potentially be a good source for

various stem cell-based therapeutic applications. DSCs can also be
reprogrammed into induced pluripotent stem cells (iPSCs) that
possess a tremendous wide range of biomedical applications
[11–13].
There are at least six types of DSCs from the human teeth that
have been isolated and characterized [1–3]. The first stem cells of
the dental mesoderm were isolated from the dental pulp of the
human permanent teeth, termed postnatal dental pulp stem cells
(DPSCs) [14]. Subsequently, dental stem cells were isolated from
the dental pulp of deciduous teeth [stem cells from human exfo-
liated deciduous teeth (SHED)] [15] and the periodontal ligament
[periodontal ligament stem cells (PDLSCs)] [16]. Stem cells were
also isolated from the apical papilla [stem cells from apical papilla
(SCAP)] [17] and the dental follicle [dental follicle stem cells
(DFSCs)] [18]. Here in this chapter, we will focus on describing
the material and methods established in our laboratory on isolation
and characterization of DPSCs, PDLSCs, SCAP, and SHED
in vitro and in vivo.
2 Materials
The following materials are needed for the isolation and culturing
of DSCs.
2.1 Tooth Collection α-Minimal essential medium (α-MEM) supplemented with 3

Medium concentrated antibiotic/antimycotic (1 concentration for each
component is 100 units/mL penicillin G, 100 μg/mL streptomy-
cin, and 0.25 μg/mL Fungizone).
2.2 Cell Isolation Tissue digestion solution: 3 mg/mL collagenase type I and 4 mg/
Enzymes (Digestion mL dispase. Prepare fresh to get the best result, or store in 80 C
Buffer) in aliquots.
2.3 Basic Cell α-MEM (up to 500 mL volume) supplemented with:

Culture Medium
(a) Fetal bovine serum (FBS) (10–20% of the total volume of α-
for DSCs
MEM)
(b) 100 μM L-ascorbic acid
(c) 2 mM L-glutamine
(d) 100 units/mL Pen-Strep/0.25 μg/mL Fungizone solution
2.4 Reagents to Trypsin-EDTA (porcine trypsin, 0.25%, EDTA, 2.2 mM, in PBS)
Detach Cells for
Passaging
Isolation, Characterization and Expansion of Dental Mesenchymal Stem Cells 61
2.5 Cryopreservation 90% FBS

Medium 10% Dimethyl sulfoxide (DMSO)
2.6 Osteogenic α-MEM supplemented with:

Differentiation Medium
10% FBS
10 nM dexamethasone
10 mM β-glycerophosphate
50 μg/mL ascorbate phosphate
10 nM 1,25-dihydroxyvitamin D3
2.7 Adipogenic α-MEM supplemented with:

Differentiation Medium
10% FBS
1 μM dexamethasone
1 μg/mL insulin
0.5 mM 3-isobutyl-1-methylxanthine (IBMX)
2.8 Chondrogenic High-glucose DMEM supplemented with:

Differentiation Medium 100 nmol/L dexamethasone
50 μg/mL ascorbic acid-2-phosphate
100 μg/ml sodium pyruvate
40 μg/mL L-proline
10 ng/mL recombinant human transforming growth factor-β3
(TGF-β3)
50 mg/mL ITS-premix stock (BD Biosciences)
2.9 Neurogenic 1. Neurobasal A supplemented with:

Differentiation Medium (a) B27 supplement
(b) 20 ng/mL epidermal growth factor (EGF)
(c) 40 ng/mL fibroblast growth factor (FGF)
2. α-MEM supplemented with:
(a) 10 ng/mL bFGF
(b) 10 μM forskolin (Sigma)
(c) 25 mM KCl
(d) 2 mM valproic acid
(e) 5 μg/mL insulin
3 Methods
This section describes laboratory protocols of DSC isolation, cul-

turing, expansion, and basic stem cell characterization.
3.1 Tooth Tissue 1. Collect freshly extracted teeth from the dental clinic in sterile
Isolation tubes filled with the tooth collection medium, and transport to
the laboratory for processing.
2. From this point on, all procedures are performed in the tissue
culture hood.
3. The teeth will be further washed and stored at 4 C (see
Note 1).
4. Carefully collect the desired tissue for stem cell isolation. Begin
with tissues on the tooth surface, periodontal ligament, and
apical papilla of immature teeth. Note: the large soft tissue at
the cervical region is the coronal follicle merging into the
gingiva (Fig. 1a, b).
5. To remove pulp tissue, the tooth needs to be split open (see
Note 2). The pulp tissue is then removed with a pair of cotton
pliers or endodontic files (Fig. 1c).
6. Removed tissues are placed in microcentrifuge tubes contain-
ing the Digestion Buffer.
3.2 Cell Isolation 1. Isolation of DSCs follows a protocol described previously in

the literature [14, 17, 19, 20].
2. Wash the collected tissue three times with the tissue collection
medium.
3. Cut the tissue into fragments as small as possible
(<1 1 1 mm) with scissors, while the tissue is in the
microcentrifuge tube with Digestion Buffer in it.
4. Tissues are then incubated for 30–60 min at 37 C. For more
effective digestion, place the tubes into a thermomixer set to
37 C, and shake during digestion (see Note 3).
5. At the end of the digestion, the tissue pieces become very loose
and disintegrated.
6. The remaining loose tissue fragments are passed through a
70 μm cell strainer to obtain maximum yield of single cell
suspension.
7. Centrifuge the cell suspension.
8. Remove the supernatant and resuspend cells in the culture
medium.
9. Combine all the collection of resuspended cells, and plate them
into culture plates.
Fig. 1 Freshly extracted human teeth for DSC isolation. (a) Extracted third molars with immature apecies
carrying the apical papilla. No clinical information is available. Judging by the presence and location of the
follicle, the tooth on the left is likely fully erupted leaving behind a half piece of coronal follicle. The tooth on
the right has emerged from the alveolar bone but partially covered by the follicle. (b) Extracted third molars
with more mature apex. The apical papilla is no longer present or too small to isolate. The tooth on the left was
fully erupted showing the small amount of attached gingiva at the cervical region. The tooth on the right still
carried a piece of the follicle indicating likely not fully erupted. The coronal follicle is normally softer than the
gingiva tissue. Unless the tooth is completely impacted in the bone, the follicle is likely mixed with adjacent
gingival tissues. They are not well-studied. (c) After external tissues are removed, the tooth was cut open to
extract the pulp.
10. Cells are seeded into 12-well plates, cultured with culture
media, and maintained in 5% CO2 at 37 C (see Note 4).
11. In the first 3–4 days, no refreshing of the medium to allow as
many cells to settle and attach.
12. Add fresh medium after 3 days, and you may replace old
medium at day 7.
13. Colony-forming units of fibroblastic cells (CFU-F) are nor-
mally observed within 1–2 weeks after cell seeding.
14. With 14 days of culture, aggregates of 50 cells/colonies of
cells should appear (Fig. 2a–c). This is considered as cells at
passage 0. If no cells are seen after 2–3 weeks, the isolation was
not successful.
Fig. 2 Colony formation of DSCs at passage 0. (a, b) DPSCs and SCAP were from an 18-year-old female donor,
showing various sizes of colonies with different cell densities after 10 days of plating. (c) PDLSCs were
isolated from a 25-year-old female donor, showing also various sizes and densities of cell colonies after
7 days of seeding. Quite a number of epithelial islands (yellow arrow) can be observed among PDLSCs. These
epithelial cells are from the epithelial cell rests of Malassez. (d) PDLSCs formed colonies and stained with
0.1% toluidine blue (left, 10 CM dish, at passage 3, donor 17-year-old female; right two images at passage 1,
donors left and right, 25-year-old and 26-year-old, respectively, no gender info). Scale bar: top left image,
200 μm for all bright field images; bottom panel, 1 mm for two images on the right
15. The success rate, with no contamination, is around 80–90% of

isolation for trained workers. Observation of contamination
may occur any time from day 2 to more than a week after cell
seeding (see Note 5).
16. At passage 0, the cultures may be precipitated with numerous
small debris from the digested tissues. The debris further
breaks down over time (see Note 6).
17. Variation between donors is considerable (see Note 7).
3.3 Cell Expansion 1. After 7 days, replenish the culture medium every 2–3 days.
When cells reach ~70% confluence, they should be passaged
at 1:3 ratio as follows:
(a) Wash the cell monolayer with PBS.
(b) Add optimal amount of trypsin-EDTA to the culture
plate.
(c) Place the plate at 37 C for 5 min, and inspect the mono-
layer under magnification. The adherent cells should be
detached from the culture plate.
(d) After cells detach and become a single cell suspension, add
fresh medium and dispense them into new plates.
3.4 Cell 1. Isolated cells at any passage may be cryopreserved and later
Cryopreservation thawed and expanded for experimentation.
2. Detach the cells as described above and add fresh culture
medium. Centrifuge the cell suspension.
3. Resuspend the cell pellet in freezing medium and aliquot into
cryovials.
4. Place the cryovials into cell freezing container, and store at
80 C overnight.
5. Transfer the vials to liquid nitrogen for longer storage (see
Note 8).
3.5 Cell Thawing 1. Place the cryovial from the liquid nitrogen in the 37 C water
bath to thaw with agitation.
2. After ~70% of the frozen liquid in the vial is thawed, spray the
vial with 70% EtOH to clean the vial and work in the tissue
culture hood.
3. Open the vial and add prewarmed culture medium dropwise
into the vial.
4. The following two steps are optional:
5. Transfer the cell suspension to a tube filled with more culture
media and centrifuge. Discard the supernatant.
6. Add fresh culture medium, and equally aliquot the suspension
into culture plates.
7. Incubate at 37 C under 5% CO2.
8. The next day, wash the cells with PBS (optional), and add fresh
culture medium.
9. Expand cells by passaging at ~70–80% confluence at 1:3 ratio
(see Note 9). Observe the morphology and growth rate of the
cells carefully; cells turning larger and growing slower are com-
mon signs of aging.
3.6 Colony-Forming One cardinal sign of DSCs is to form CFU even at higher passages
Unit (CFU) Assay such as >3.
1. Seed cells at low density (~60 cells/cm2), and allow each cell to
have space for colony formation.
2. After ~1 week, observe the CFU-Fs. Aspirate the medium, and
wash the dishes three times with PBS, and fix cells with 10%
neutral buffered formalin for ~3–5 min.
3. Add 0.5% crystal violet or 0.1% toluidine blue solution to fixed
cells, and incubate for 5–10 min at room temperature.
4. Aspirate the stain, and then wash with excess distilled water
until the background is clear.
5. Count the number of colonies for each dish, determine the
mean, and calculate the plating efficiency or “CFU potential”
(% CFU formed relative to inoculum). A good culture of MSCs
typically has a CFU potential of over 40% [21] (Fig. 2d).
3.7 Marker In 2006, the International Society for Cellular Therapy (ISCT)
Expression proposed markers that define human multipotent MSCs, CD146,
CD105, CD73, and CD90, and lack the expression of CD45,
CD34, CD14 or CD11b, CD79α or CD19, and HLA-DR surface
molecules [22, 23]. For DSCs, CD146, CD105, CD73, and CD90
are expressed [1]. Marker expression can be detected by flow cyto-
metry or immunocytofluorescence.
3.7.1 Flow Cytometry 1. For direct staining of cell surface antigens, subconfluent cells
are harvested and washed, resuspended in staining buffer
(PBS + 0.1% FBS), and incubated with conjugated FITC or
Alexa Fluor antibodies for 30 min at 4 C according to the
manufacturer’s recommendations.
2. For intracellular antigens, single cell suspensions are first fixed
in 4% paraformaldehyde in PBS for 10 min, washed with flow
buffer, permeabilized in 0.1% Triton X-100 for 10 min, washed
with flow buffer, and stained as above.
3. Cell suspensions are then washed twice with flow buffer and
resuspended in flow buffer for analysis on a flow cytometer
(FACSCalibur, BD Biosciences) using the CellQuest ProTM
software (BD Biosciences). Figure 3 shows a typical flow cyto-
metry analysis.
3.7.2 Immunocytofluore- 1. Cells grown in chamber glass slides (eight wells) or in culture
scence plates are washed and fixed with 100% ice-cold methanol for
7–10 min.
2. After PBS washing, cells are blocked with 5% goat serum in PBS
or in blocking buffer (32.5 mM NaCl, 3.3 mM Na2HPO4,
0.76 mM KH2PO4, 1.9 mM NaN3, 0.1% [w/v] bovine serum
albumin (BSA), 0.2% [v/v] Triton X-100, 0.05% [v/v] Tween
20, and 5% goat serum) for 30 min.
105 Ctrl CD73 CD90
105
105
104
104
104
Count
103
103
103
102
102
102
102 103 104 105 102 103 104 105 102 103 104 105
CD105 CD146 STRO-1
105
105
105
104
104
104
Count
103
103
103
102
102
102
102 103 104 105 102 103 104 105 102 103 104 105
Fig. 3 Immunophenotype analysis of human dental pulp stem cells by flow cytometry. Multiple colony-derived
cells (1 105) at passage 2 were incubated with specific monoclonal antibodies against cell surface marker
antigens CD73, CD90, CD105, CD146, and STRO-1. Percentage of positively stained cells is indicated on the
right side of the plot. Subclass-matched control antibodies were used for the controls (Ctrl). DPSCs from a
20-year-old male, tooth #17
3. The primary antibody is then added directly to cells and incu-

bated for 1 h at room temperature and washed with PBS for
three times each 5 min on a rocker.
4. After PBS wash, secondary antibody (Alexa Fluor 594 or Alexa
Fluor 488) in blocking buffer is added and incubated for 1 h at
room temperature in dark.
5. Cell nuclei are stained with 40 ,6-diamidino-2-phenylindole
dihydrochloride (DAPI) for 3 min. Images are analyzed
under a fluorescence microscope (Fig. 4).
3.8 In Vitro Multi- 1. Cells are seeded into 12-well or 24-well plates, grown to ~70%
lineage Differentiation confluence, and incubated in osteogenic differentiation
Assay medium at 37 C under 5% CO2 for 4 weeks.
3.8.1 Odonto-/ 2. Replenish the osteogenic medium every 2–3 days.
Osteogenic Differentiation 3. At the end of the culture period, cultures are rinsed twice with
PBS, fixed in 60% isopropanol for about 1 min at room tem-
perature, and washed three times with distilled water (dH2O).
Fig. 4 Immunophenotype analysis of human dental pulp stem cells by immunocytofluorescence. (a) Cell
colonies (50 cells per colony) at passage 0 were immunocytostained after 1 week of their initial seeding.
Approximately 40–90% of the colonies showed positive staining for STRO-1, CD73, CD90, CD105, or CD146.
STRO-1 fluorescence staining (red), CD staining (green), and DAPI nuclear staining (blue). (b) Immunocyto-
fluorescence staining of STRO-1, CD73, CD90, CD106, CD146, and isotype control. Subconfluent cells were all
at passage 2. Isotype control (Ctrl). Fluorescence staining (red) and DAPI nuclear staining (blue). Scale
bars ¼ (A) 200 μm for all images in except STRO-1 = 100 μm. (B) All 100 μm
4. The mineralization of extracellular matrix is stained with 1%

alizarin red S (ARS) in dH2O at room temperature for 30 min
and then washed five times with dH2O.
5. For quantification of staining, cultures are fixed with 70%
icescold ethanol for 1 hour at room temperature and then
washed three times with dH2O.
Fig. 5 Osteogenic differentiation of human DPSCs. (a) Alizarin red stain showing mineral deposits after
4 weeks of osteogenic induction. (b) Quantitative alizarin red measurements represented by fold change.
OSTEO, osteogenic induction media; CTRL, DPSC basic culture media without induction
6. ARS is extracted from the cultures by incubation of the mono-

layers in cetylpyridinium chloride (CPC) buffer (10% CPC
(w/v) in 10 mM sodium phosphate buffer) for 1 hour at
room temperature.
7. The dye is then removed, and three 200 μL aliquots of
ARS/CPC extract from each well are transferred to a 96-well
plate and quantified by absorbance measurement at 550 nm by
spectrophotometer (Bio-Rad) (Fig. 5).
3.8.2 Adipogenic 1. Cells are seeded into 12-well or 24-well plates, grown to sub-
Differentiation confluence, and incubated in adipogenic differentiation
medium at 37 C under 5% CO2 for 8 weeks.
2. Replenish the adipogenic medium every 2–3 days.
3. At the end of the culture period, cultures are rinsed twice with
PBS, fixed in 10% formalin for 60 min at room temperature,
and washed three times with dH2O.
4. Lipid droplets are stained for 5 minutes at room temperature
with 0.18% oil red O (ORO) working solution (ORO 0.3%
(w/v) in isopropyl alcohol (three parts), and dH2O (two parts)
filter through a 70 μm cell strainer or filter paper. Freshly made,
stable for 2 h).
5. The cultures are then washed three times with dH2O, and
images are taken by a microscope.
6. For quantitative analysis of the staining, ORO stain was
extracted with 60% isopropyl alcohol for 10 min at room
temperature. Three aliquots (200 μL) of the extracted oil red
O were transferred to a 96-well plate and quantified by absor-
bance measurement at 540 nm by spectrophotometer
(Bio-Rad) (Fig. 6).
Fig. 6 Adipogenic differentiation of human DPSCs. Oil red stain showing oil droplet-filled adipocyte-like cells
after 8 weeks of adipogenic induction. CTRL, DPSC basic culture media without induction; ADIPO, adipogenic
induction media. Scale bar ¼ 150 μm for the left and center image; 75 μm for the right image
3.8.3 Chondrogenic 1. Cells are seeded into 24-well plates (monolayer assay) or into
Differentiation 200 μL microfuge tubes and centrifuged down to form cell
pellets (pellet culture).
2. Pellet culture can also be performed in a 96-well plate
(V-bottom well).
3. Cell cultures in plates or in tubes are treated for 3 weeks with
chondrogenic differentiation medium, and the medium
changed every 3 days.
4. Chondrogenic cell cultures or pellets are fixed in 10% neutral
buffered formalin for 60 min.
5. Cultures are washed with water and incubated overnight with
Alcian Blue staining solution prepared by dissolving 10 mg
Alcian Blue 8 GX in 100 mL solution with 60% ethanol and
40% acetic acid (v/v) (pH 3.0). Alcian blue is to detect sulfated
proteoglycans.
6. The cells or pellets are washed with destaining solution with
60% ethanol and 40% acetic acid (v/v) before being analyzed
(Fig. 7).
7. For histology staining, pellets were first embedded in agarose
gel and then fixed and processed for paraffin embedment and
sectioning. Sections are deparaffinized and stained with Alcian
Blue for 30 min, followed by washing in running tap water for
2 min.
8. Sections are then counterstained in nuclear fast red solution for
5 min.
9. Wash in running tap water for 1 min, and dehydrate through
95% alcohol, two changes of absolute alcohol, 3 min each.
Subsequently, sections are mounted in resinous mounting
medium (Fig. 7). Note the histological view of Alcian Blue
stain of DSCs after chondrogenesis seldom shows typical chon-
drocyte morphology, although samples often show Alcian
Blue-positive staining.
Fig. 7 Chondrogenic differentiation of PDLSCs. PDLSCs at passages 2–3 were seeded, and after reaching
~70% confluence, they were stimulated under the chondrogenic condition. (a) Chondrogenic stimulation for
21 days and stained with Alcian Blue. Cells cultured in 24-well plates (2 104 cells/well) as monolayers; note
the cell contraction into spheres after stimulation. (b) Cells cultured as pellets in 200 μL microfuge tubes
(5 104 cells/tube) or in 96-well plate (V-bottom). Representative data showing Alcian Blue-stained pellet
without histological processing (left) or after histological processing (right). Scale bar ¼ 100 μm (left), 300 μm
(right)
3.8.4 Neurogenic Medium 1

Differentiation
1. Cells at subconfluence in chamber slides or in 12- or 24-well
plates are stimulated by the neurogenic induction medium for
4 weeks with the medium refreshed every 3 days.
2. Cells are monitored continually after neural induction for mor-
phological changes.
3. Cultures are then analyzed by immunocytofluorescence or
RT-qPCR for the expression of the neural cell markers
(Fig. 8a).
Fig. 8 Neurogenic differentiation of DSCs. Cells at passages 2–3 were seeded, and after reaching ~70%
confluence, they were stimulated under the neurogenic condition and induced for 4 weeks (DPSCs) or for
2 weeks (PDLSCs) and stained for βIII tubulin (red or green fluorescence) with DAPI nuclear counterstain. Scale
bar ¼ all four images, 20 μm
Medium 2
1. Cells receive preneural induction in α-MEM medium contain-
ing 10% FBS and 10 ng/mL bFGF for 24 h [24].
2. Subsequently, the medium was removed, cells washed with
PBS, and the medium added for up to 35-day incubation
period with the medium refreshed every 3 days (Fig. 8b).
3.9 In Vivo Tissue 1. Over-confluent cells (2 days after confluence) are harvested and
Formation mixed with hydroxyapatite and tricalcium phosphate
(HA/TCP) granules (20 mg, Berkeley Advanced Biomaterials
3.9.1 Cells and
Inc., Berkeley, CA) in a 2 mL tube.
Hydroxyapatite/Tricalcium
Phosphate Mixture 2. Tubes are incubated with rotation for 90 min at 37 C.
3. The mixture (2–3 106 cells/40 mg HA/TCP) is pelleted.
Each mixture measures ~ 5 5 4 mm.
4. The mixtures are transplanted into the back of a NOD.CB17-
Prkdcscid/J mouse (male or female, ~7 week-old) (Jackson
Labs., Bar Harbor, Maine).
5. Each pelleted mixture is carefully transferred using a periodon-

tal flap elevator instrument into the subcutaneous space (cre-
ated by incision) of the mouse.
6. Two to three months after implantation, the mice are eutha-
nized and transplants removed for histological analysis (Fig. 9).
4 Notes
1. For best result, cell isolation should be performed within 24 h

of tooth collection.
2. It is normally done by cutting the crown off beginning with
making a deep groove all around the cervical region using
dental burs, followed by using a sterile flat blade instrument
to pry open the tooth.
3. Optional: For preventing damage to the early released cells by
the Digestion Buffer, the supernatant is collected every 15 min,
and the cells spun down followed by resuspending in the
culture medium. The undigested tissue is placed back to the
Digestion Buffer for further digest until finished.
4. Plating cells into a 12-well plate is to prevent complete con-
tamination of all the collected cells. Our experience indicates
that occasionally we will find a few wells contaminated, while
the remaining wells not.
5. Cell isolation from pulp tissue tends to have the least chance of
contamination as it is protected in the tooth.
6. At times, it is difficult to differentiate contamination vs. debris.
Careful observation at high magnification (40 objective)
under the microscope is required.
7. Cells isolated from younger donors tend to yield more robust
cells.
8. Frozen cells should not be stored in 80 C for more than
1–2 months to reduce possibility of lower survival when
thawed.
9. Note: Depending on the condition of the cells, 1:5 ratio for
passaging may be used. The most important principle is the
population doubling (PD). Keep the cells at low PD for experi-
mentation. Higher passages of stem cells will yield more differ-
entiated cells and less stem cells/progenitor cells.
For more information especially data presentation, please refer
to the following References (4, 19, 25–29). The method is suited
for isolation of any dental/oral stem cells including stem cells from
the primary teeth, although we do not show data from cells of such
teeth.
Fig. 9 In vivo tissue formation in mice. Cells and hydroxyapatite/tricalcium phosphate (HA) mixture formed
mineral tissue (yellow arrows) around the HA granules after 2–3 months of transplantation. In between is the
soft fibrous tissue. (a, d) Lower magnification views of the entire tissue mass resected from mice containing
the mixture with new tissue formation. (b–f) Higher magnification views of the tissue. DPSCs/HA formed pulp-
dentin complex-like tissues; PDLSCs/HA formed PDL-cementum or bone-like tissues. Scale bars ¼ (a,d)
1 mm, (b,e) 500 μm, (c,f) 100 μm
Acknowledgments
This work was supported in part by a grant from the US National

Institutes of Health R01 DE019156 (G.T.-J.H.), by an Endodon-
tic Research Grant from American Association of Endodontists
Foundation (G.T.-J.H.), and by a Research Fund from the Univer-
sity of Tennessee Health Science Center. The authors wish to thank
Drs. Philippe Guathier, Zongdong Yu, and Yaa Owusu for their
work on PDLSCs.
References
1. Huang GTJ, Gronthos S, Shi S (2009) Mesen- neuro-regenerative mechanisms. J Clin Invest
chymal stem cells derived from dental 122(1):80–90
tissues vs. those from other sources: their biol- 9. Gay IC, Chen S, MacDougall M (2007) Isola-
ogy and role in regenerative medicine. J Dent tion and characterization of multipotent
Res 88(9):792–806 human periodontal ligament stem cells.
2. Morsczeck C, Huang GT-J, Shi S (2013) Stem Orthod Craniofac Res 10(3):149–160
and progenitor cells of dental and gingival tis- 10. Choi S, Cho TJ, Kwon SK, Lee G, Cho J
sue origin. In: Huang GT-J, Thesleff I (eds) (2013) Chondrogenesis of periodontal liga-
Stem cells, craniofacial development and ment stem cells by transforming growth
regeneration, 1st edn. Wiley-Blackwell, Hobo- factor-beta3 and bone morphogenetic
ken, NJ, pp 285–302 protein-6 in a normal healthy impacted third
3. Huang GTJ, Garcia-Godoy F (2017) Stem molar. Int J Oral Sci 5(1):7–13
cells and dental tissue reconstruction. In: 11. Yan X, Qin H, Qu C, Tuan RS, Shi S, Huang
Spencer P, Misra A (eds) Material-tissue inter- GT (2010) iPS cells reprogrammed from
facial phenomena, 1st edn. Elsevier, Woodhead human mesenchymal-like stem/progenitor
Publishing, Sawston, Cambridge cells of dental tissue origin. Stem Cells Dev 19
4. Huang GTJ, Yamaza T, Shea LD, Djouad F, (4):469–480
Kuhn NZ, Tuan RS et al (2009) Stem/progen- 12. Huang GT-J (2010) Induced pluripotent stem
itor cell-mediated de novo regeneration of den- cells—a new foundation in medicine. J Exp
tal pulp with newly deposited continuous layer Clin Med 2(5):202–217
of dentin in an in vivo model. Tissue Eng Part 13. Huang GTJ, El Ayachi I, Zou X-Y (2017)
A 16(2):605–615 Induced pluripotent stem cell technologies for
5. Nakashima M, Huang GT-J (2013) Pulp and tissue engineering. In: Waddington RJ, Sloan
dentin regeneration. In: Huang GT-J, Thesleff AJ (eds) Tissue engineering and regeneration
I (eds) Stem cells in craniofacial development in dentistry- current strategies. John Wiley &
and regeneration, 1st edn. Wiley-Blackwell, Sons, Ltd, West Sussex, UK
Hoboken, NJ, pp 461–484 14. Gronthos S, Mankani M, Brahim J, Robey PG,
6. Ishizaka R, Hayashi Y, Iohara K, Sugiyama M, Shi S (2000) Postnatal human dental pulp stem
Murakami M, Yamamoto T et al (2013) Stim- cells (DPSCs) in vitro and in vivo. Proc Natl
ulation of angiogenesis, neurogenesis and Acad Sci U S A 97(25):13625–13630
regeneration by side population cells from den- 15. Miura M, Gronthos S, Zhao M, Lu B, Fisher
tal pulp. Biomaterials 34(8):1888–1897 LW, Robey PG et al (2003) SHED: stem cells
7. Iohara K, Zheng L, Wake H, Ito M, from human exfoliated deciduous teeth. Proc
Nabekura J, Wakita H et al (2008) A novel Natl Acad Sci U S A 100(10):5807–5812
stem cell source for vasculogenesis in ischemia: 16. Seo BM, Miura M, Gronthos S, Bartold PM,
subfraction of side population cells from dental Batouli S, Brahim J et al (2004) Investigation
pulp. Stem Cells 26(9):2408–2418 of multipotent postnatal stem cells from
8. Sakai K, Yamamoto A, Matsubara K, human periodontal ligament. Lancet 364
Nakamura S, Naruse M, Yamagata M et al (9429):149–155
(2012) Human dental pulp-derived stem cells 17. Sonoyama W, Liu Y, Yamaza T, Tuan RS,
promote locomotor recovery after complete Wang S, Shi S et al (2008) Characterization of
transection of the rat spinal cord by multiple the apical papilla and its residing stem cells
from human immature permanent teeth: a pilot International Society for Cellular Therapy posi-
study. J Endod 34(2):166–171 tion statement. Cytotherapy 8(4):315–317
18. Morsczeck C, Vollner F, Saugspier M, 24. Ebrahimi B, Yaghoobi MM, Kamal-abadi AM,
Brandl C, Reichert TE, Driemel O et al Raoof M (2011) Human dental pulp stem cells
(2010) Comparison of human dental follicle express many pluripotency regulators and dif-
cells (DFCs) and stem cells from human exfo- ferentiate into neuronal cells. Neural Regen
liated deciduous teeth (SHED) after neural Res 6(34):2666–2672
differentiation in vitro. Clin Oral Investig 14 25. Alongi DJ, Yamaza T, Song Y, Fouad AF,
(4):433–440 Romberg EE, Shi S et al (2010) Stem/progen-
19. Zhu X, Liu J, Yu Z, Chen CA, Aksel H, Azim itor cells from inflamed human dental pulp
AA, Huang GTJ (2018) A miniature swine retain tissue regeneration potential. Regen
model for stem cell-based de novo regenera- Med 5(4):617–631
tion of dental pulp and dentin-like tissue. Tis- 26. Gauthier P, Yu Z, Tran QT, Bhatti FU, Zhu X,
sue Eng Part C Methods 24(2):108–120 Huang GT (2017) Cementogenic genes in
20. Sonoyama W, Liu Y, Fang D, Yamaza T, Seo human periodontal ligament stem cells are
BM, Zhang C et al (2006) Mesenchymal stem downregulated in response to osteogenic stim-
cell-mediated functional tooth regeneration in ulation while upregulated by vitamin C treat-
swine. PLoS One 1:e79 ment. Cell Tissue Res 368(1):79–92
21. Smith JR, Pochampally R, Perry A, Hsu SC, 27. Yu Z, Gauthier P, Tran QT, El Ayachi I, Bhatti
Prockop DJ (2004) Isolation of a highly clono- F-U-R, Bahabri R et al (2015) Differential
genic and multipotential subfraction of adult properties of human ALP+ periodontal liga-
stem cells from bone marrow stroma. Stem ment stem cells vs their ALP- counterparts. J
Cells 22(5):823–831 Stem Cell Res Ther 5(7):292
22. Horwitz EM, Le Blanc K, Dominici M, 28. Huang GT, Sonoyama W, Liu Y, Liu H,
Mueller I, Slaper-Cortenbach I, Marini FC Wang S, Shi S (2008) The hidden treasure in
et al (2005) Clarification of the nomenclature apical papilla: the potential role in pulp/dentin
for MSC: The International Society for Cellular regeneration and bioroot engineering. J Endod
Therapy position statement. Cytotherapy 7 34(6):645–651
(5):393–395 29. Al-Habib M, Yu Z, Huang GT (2013) Small
23. Dominici M, Le Blanc K, Mueller I, Slaper- molecules affect human dental pulp stem cell
Cortenbach I, Marini FC, Krause DS et al properties via multiple signaling pathways.
(2006) Minimal criteria for defining multipo- Stem Cells Dev 22(17):2402–2413
tent mesenchymal stromal cells. The
Chapter 8
Phenotypic Identification of Dental Pulp Mesenchymal

Stem/Stromal Cells Subpopulations with Multiparametric
Flow Cytometry
Maxime Ducret, Jean-Christophe Farges, Marielle Pasdeloup,
Emeline Perrier-Groult, Andreas Mueller, Frédéric Mallein-Gerin,
and Hugo Fabre
Abstract
Dental pulp (DP) is a specialized, highly vascularized, and innervated connective tissue mainly composed of
undifferentiated mesenchymal cells, fibroblasts, and highly differentiated dentin-forming odontoblasts.
Undifferentiated mesenchymal cells include stem/stromal cell populations usually called dental pulp
mesenchymal stem cells (DP-MSCs) which differ in their self-renewal properties, lineage commitment,
and differentiation capabilities. Analysis of surface antigens has been largely used to precisely identify these
DP-MSC populations. However, a major difficulty is that these antigens are actually not specific for MSCs.
Most of the markers used are indeed shared by other cell populations such as progenitor cells, mature
fibroblasts, and/or perivascular cells. Accordingly, the detection of only one of these markers in a cell
population is clearly insufficient to determine its stemness. Recent data reported that multiparametric flow
cytometry, by allowing for the detection of several molecules on the surface of one single cell, is a powerful
tool to elucidate the phenotype of a cell population both in vivo and in vitro. So far, DP-MSC populations
have been characterized mainly based on the isolated expression of molecules known to be expressed by
stem cells, such as Stro-1 antigen, melanoma cell adhesion molecule MCAM/CD146, low-affinity nerve
growth factor receptor p75NTR/CD271, and the mesenchymal stem cell antigen MSCA-1. Using multi-
parametric flow cytometry, we recently showed that human DP-MSCs are indeed phenotypically heteroge-
neous and form several populations.
The present paper describes the multiparametric flow cytometry protocol we routinely use for character-
izing DP-MSCs. The description includes the design of the antibody panel and explains the selection of the
different parameters related to the data quality control.
Key words Dental pulp, Mesenchymal stem/stromal cells, Flow cytometry
1 Introduction
The dental pulp (DP) is a specialized connective tissue which is

responsible for tooth vitality, pain sensation, immune defense, and
repair/regeneration. Many studies have shown that it contains
77
78 Maxime Ducret et al.
various populations of mesenchymal-type cells, some of which

demonstrate stem cell properties such as expression of specific
markers, high growth potential, and multiple differentiation capa-
cities. Studies have established that there exists considerable het-
erogeneity among DP mesenchymal stem cells (DP-MSCs)
according to their origin, tooth development stage, or age of the
patient [1–4]. This heterogeneity was found to be increased upon
culture depending on culture time and medium used and number
of passages.
In 2006, the International Society for Cellular Therapy (ISCT)
proposed minimal requirements to define human cells as MSCs:
adherence to plastic, multipotent differentiation potential and spe-
cific surface antigen expression [5]. However, it is today acknowl-
edged that the surface markers initially proposed by the ISCT for
the positive characterization of MSCs (mainly CD73, CD90, and
CD105) are shared by other populations such as mesenchymal
progenitor cells, mature fibroblasts, or perivascular cells [6–8]. Spe-
cific surface antigens or characteristics are still missing or inconsis-
tent. To solve this problem, the co-localization of several molecular
markers on the surface of an individual cell was recently proposed.
Multiparametric flow cytometry is a technology which allows
for the simultaneous detection of several surface antigens by using
specific antibody cocktails. It appears to be a powerful tool to
elucidate the complex surface phenotype of MSCs. However, a
major difficulty of this technology is the ability to reliably discrimi-
nate between antigen-positive and antigen-negative cells and to
accurately characterize the population of positive cells. Thus, the
correct exploitation of this technology should combine the use of
several types of controls to ensure the accuracy of the test samples’
analysis. This paper describes the design of an antibody panel for
multiparametric flow cytometry analysis of DP-MSCs and explains
the selection of the different parameters related to the data quality
control according to international guidelines.
2 Materials
2.1 Tissue Collection 1. 50 mL Falcon tubes (Corning Inc., Corning, NY, USA).
2. Phosphate-buffered saline (PBS) w/o Ca2+ and Mg2+ (Lonza,
Basel, Switzerland).
3. Penicillin/streptomycin (Lonza).
2.2 DP-MSC Isolation 1. Collagenase type I (Calbiochem, San Diego, CA, USA).
and Culture 2. Dispase (Roche Diagnostics, Meylan, France).
3. 100-μm nylon mesh filters (Corning Inc.).
Multiparametric Flow Cytometry and Dental Pulp Mesenchymal Stem/Stromal. . . 79
4. 6-well plates (Becton Dickinson [BD], Le Pont-de-Claix,

France).
5. T75 (75 cm2) culture flasks (Corning Inc.).
6. Human placental collagens I and III (ABCell-Bio, Paris,
France).
7. SPE-IV medium (ABCell-Bio).
8. Penicillin/streptomycin (Lonza).
9. Xeno-free recombinant protease TrypLE Select 1, (Life Tech-
nologies, Saint Aubin, France).
10. 15 mL Falcon tubes (BD).
12. PBS w/o Ca2+ and Mg2+ (Lonza).
2.3 Flow Cytometry 1. Staining buffer (BD).

Analysis 2. VersaComp compensation beads (Beckman Coulter, Brea,
CA, USA).
3. Round-bottom flow cytometry tubes (BD).
6. 100-μm nylon mesh filters (Corning Inc.).
7. Cytofix/Cytoperm (BD).
8. TruStain FcX Fc receptor blocking solution (BioLegend, San
Diego, CA, USA).
9. Cytofix/Cytoperm (BD).
10. Cytometer Setup and Tracking (CS&T) beads (BD).
2.4 Antigens and 1. The staining panel is designed by using six fluorochrome-
Conjugates conjugated antibodies (Table 1); the nucleic acid dye 7-AAD
(7-aminoactinomycin D, BD-Biosciences) is used for the exclu-
sion of nonviable cells.
2. Spectral overlap is minimized by choosing combinations of
fluorochromes that have little to no overlap with each other
or by choosing multiple fluorochrome-specific independent
excitation sources.
2.5 Equipment 1. BD Lyse Wash Assistant (BD).

2. BD FACSCanto II (BD) equipped with three lasers (violet
[405 nm], blue [488 nm], and red [633 nm]).
3. FlowJo vX (Tree star Inc., San Carlos, CA, USA).
Table 1
Fluorochrome-conjugated monoclonal antibodies used for immunophenotypic analysis
Target Fluorochrome Clone Manufacturer Reference Isotype Reference

CD56 BV510 HCD56 Biolegend 318340 Mouse IgG1, κ 562946
CD146 AF488 SHM-57 Biolegend 342008 Mouse IgG2a, κ 400233
Stro-1 PE IgM, I Santa Cruz sc-47733 PE Mouse IgM, I sc-2870
CD271 PE-Cy7 C40-1457 BD-Biosciences 562852 Mouse IgG1, κ 557646
MSCA-1 APC W8B2 Biolegend 327308 Mouse IgG1, κ 400120
CD31 APC-Cy7 WM59 Biolegend 303120 Mouse IgG1, κ 400128
3 Methods
3.1 Dental Pulp 1. Select only disease-free donors aged 13–17 years.
(DP) Collection 2. Extract healthy impacted human third molars.
3. Place teeth in a 50 mL Falcon tubes (one donor per tube)
containing PBS w/o Ca2+ and Mg2+ containing 1% penicil-
lin/streptomycin (¼ 100 IU/mL penicillin/100 μg/mL strep-
tomycin), and transport them to the laboratory within 12 h
(Fig. 1a).
4. Extirpate gently, aseptically the DP from the pulp chamber
with fine tweezers (Fig. 1b).
5. Remove the apical part of the DP with a scalpel to prevent
contamination by dental apical papilla and periodontal cells.
6. Rinse the DP in PBS containing 1% penicillin/streptomycin,
and cut it with a scalpel into 0.5–2 mm3 fragments (Fig. 1c).
3.2 Isolation of DP 1. Collect DP fragments in 15 mL Falcon tubes.

Cells for In Vivo 2. Digest fragments in a mixture of 3 mg/mL collagenase type I
Analysis and 4 mg/mL dispase for 1 h at 37 C.
3. Wash the cell suspension twice by centrifugation at 500 g for
5 min, and resuspend the pellet in 20 mL of sterile PBS. Filter it
through a 100 μm nylon mesh filter, centrifuge cells at 500 g
for 5 min, and resuspend them in sterile staining buffer.
3.3 Isolation and 1. Precoat wells of 6-well plates with an equal mixture of human
Expansion of DP Cells collagens I and III at a final concentration of 0.5 μg/cm2.
for In Vitro Analysis 2. Seed four DP fragments (¼ explants) of the same tooth in each
well (Fig. 1d).
3. Cover explants with serum-free medium SPE-IV supplemented
with 1% penicillin/streptomycin.
Fig. 1 Standardized isolation process of DP-MSCs. Human third molars were collected and transported in
PBS-/antibiotics-containing Falcon tubes (a). The DP was gently extirpated from the tooth (b) and cut into
small fragments (explants) (c). DP explants were seeded onto collagen-precoated 6-well plates (d). Cells
started to grow from the DP explants after 3 to 5 days of culture (e). During expansion, DP cells exhibited a
fibroblast-like morphology (f)
4. Incubate plates at 37 C in a 5% CO2 atmosphere. Cells start to

grow from the DP explants after 3–5 days (Fig. 1e). Change
culture medium twice a week.
Table 2
Fluorescence Minus One (FMO) isotype control strategy
Control tubes
Tube 1 Tube 2 Tube 3 Tube 4 Tube 5 Tube 6 Tube 7 Panel

Isotype CD56 CD56 CD56 CD56 CD56 CD56 CD56
CD146 Isotype CD146 CD146 CD146 CD146 CD146 CD146
Stro-1 Stro-1 Isotype Stro-1 Stro-1 Stro-1 Stro-1 Stro-1
CD271 CD271 CD271 Isotype CD271 CD271 CD271 CD271
MSCA-1 MSCA-1 MSCA-1 MSCA-1 Isotype MSCA-1 MSCA-1 MSCA-1
CD31 CD31 CD31 CD31 CD31 Isotype CD31 CD31
7-AAD 7-AAD 7-AAD 7-AAD 7-AAD 7-AAD Empty 7-AAD
5. After about 2 weeks of culture, cells become subconfluent.

Detach them with TrypLE, seed 5.103 cells/cm2 into T75
flasks, and culture them for 1 week to obtain a sufficient num-
ber of cells for immunophenotyping (Fig. 1f).
3.4 Flow Cytometry 1. Detach subconfluent cells with TrypLE, and wash the cell
Analysis suspension twice by centrifuging it at 500 g for 5 min.
Resuspend the pellet in 10 mL of sterile PBS w/o Ca2+ and
3.4.1 Sample Processing
Mg2+ (see 4.2.1).
2. Prepare a 1.107 cells/mL cell suspension in staining buffer, add
100 μL cell suspension to each of the eight different combina-
tions of antibodies (Table 2), and incubate for 25 min at 4 C in
the dark.
3. Wash cells with BD Lyse Wash Assistant (see 4.2.1) to maximize
cell viability and prevent cell adhesion to the staining tube.
4. Keep cell samples on ice, and analyze them within 2 h of
processing after a 10 min incubation with 7-AAD.
3.4.2 Antibody Panel 1. The type and number of lasers and detectors dictate whether
Design and Fluorochrome the optical system can excite a given fluorochrome and properly
Selection detect a given combination of fluorochromes. The design of
the optical system also impacts the detection efficiency of par-
ticular dyes, as do the instrument settings, including photo-
multiplier tube (PMT) voltages.
2. The choice of the optical filters that are used with each detector
greatly influences the effective brightness of one fluorochrome
(see 4.2.2).
3. Given the many differences in instrument configuration, it is
impossible to universally state the “best” fluorochromes to use
in combinations of four colors or more [9]. However, for a

particular cytometer, it is possible to rank available dyes accord-
ing to their brightness on that instrument (when configured
with a specified set of lasers and filters).
4. When the same antibody is conjugated to various dyes, their
stain indexes can be compared to get an idea of the relative
brightness of the dyes on a particular instrument. This assumes
that the conjugation chemistries for all of the reagents have
been optimized (see Subheadings 3.4.3, 3.4.4, and 4.2.2).
5. The spectral ranges of most fluorochrome emissions are so
large that the emission of a particular dye is generally measured
by many detectors rather than by the one solely designated to
measure the emission peak of that dye. This overlap of emission
spectra in the various detection regions contributes to back-
ground fluorescence and can be corrected for by utilizing
spectral compensation [10, 11] (see Subheading 3.4.6).
6. Spectral overlap can be minimized by choosing combinations
of fluorochromes that have little to no overlap with each other
or by choosing multiple fluorochrome-specific, independent
excitation sources.
7. An example of panel design can be found in Table 1.
3.4.3 Antigen Selection 1. Once the fluorochromes to be used have been defined, anti-
body specificities to particular fluorochromes can be matched
to select the actual conjugates to be used.
2. The ground rule should be to consider using the brightest
available fluorochrome for relatively dimly expressed proteins
(either the protein is not abundant on the cell surface, or the
available antibodies are of low affinity) while using a dimmer
fluorochrome for an abundant protein to which antibodies
stain cells very brightly.
3. One important thing to consider is to avoid spillover from
bright cell populations into detectors requiring high sensitivity
for those populations. The most common example would be
phycoerythrin (PE) spillover in the fluorescein isothiocyanate
(FITC) detector. If possible, the FITC-stained antigen should
be moved to a fluorochrome that has less spectral overlap with
PE (such as PerCP-Cy5.5 or APC), or the PE-stained antibody
should be moved to a detector which is still relatively bright but
which does not overlap with FITC (such as allophycocyanin
[APC] or PE-Cy5).
3.4.4 Titration Procedure 1. In addition to undesirable binding (e.g., to Fc receptors),

undesirable nonspecific antibody binding can occur. This bind-
ing is of much lower affinity than specific antigen-antibody
binding and can be controlled with blocking reagents like
nonimmune serum. Nonspecific binding is usually best elimi-

nated by optimizing antibody concentration with titration
assays.
2. Each fluorescent antibody requires careful titration at optimal
PMT voltage settings to determine the antibody concentration
that produces optimal fluorescence resolution with minimal
nonspecific background staining.
3. Titration allows determining the antibody concentration which
results in the highest signal of the positive population and the
lowest signal of the negative population. An unstained cell
sample acts as a good reference to establish the degree of
nontargeted antibody binding (background) observed in the
internal negative control population of the antibody-labeled
cell sample (Fig. 2).
4. When the positive and negative populations can be easily dis-
tinguished (antibody titer), the separation between the two are
to be expressed quantitatively by means of the resolution or the
staining index. These data should be used as a reference in
antibody titration assays and to monitor other lot numbers of
the same antibody conjugate. In addition, maintaining small
volumes and high cell concentrations during cell-labeling pro-
cedures generally help preventing high background [12, 13].
3.4.5 Controls 1. Isotype controls (4.2.5) are good controls to identify staining
issues, particularly when primary and secondary antibodies are
used compared to fluorochrome-conjugated antibodies. How-
ever, isotypes do not reliably identify negative populations from
positive ones (Fig. 3) [14].
2. Isotype controls should also be carefully titrated since nonspe-
cific activity at supersaturating levels will increase total
measured binding and will skew the negative population
more positive.
3. FMO (Fluorescence Minus One) controls contain every stain
in the panel except the one controlled for in that sample
(Table 2). Since FMO controls are labeled with all the fluor-
ochromes involved except one, they show (unlike singly stained
controls) the same apparent fluorescent shift in the negative
population as the experimental sample.
4. FMO controls help determine positivity and set regions in
samples that contain multilabeled subpopulations.
5. FMO controls can be used in combination with isotype con-
trols by replacing the missing antibody in every FMO control
tube by the corresponding isotype control (Table 2). This
method enables the visualization of staining issues (isotype
control) and gating boundaries (FMO control) without the
Fig. 2 Representative titration experiment using the FITC-conjugated anti-CD146 antibody. Cells were stained
with decreasing amounts of anti-CD146-AF488 antibody, as shown and analyzed by flow cytometry. The
0.3 μg antibody (ATB) represented the minimal background to staining ratio
need for multiple tubes, thus saving time in sample preparation

while requiring fewer cells (Figs. 3 and 4).
6. Fidelity controls use a given antibody by itself (or with minimal
additional gating reagents). They compare the results with
those obtained when this antibody is present in a complete
cocktail.
–103 0 103 104 105 –103 0 103 104 105

CD271 Per-CPCy5.5 Stro-1 Alexa Fluor 647
Fig. 3 Example of isotype gating-induced error. MSCs were stained with the two markers of interest CD271
and Stro-1. When gated against isotype controls, positivity is overestimated for the CD271 marker (a, isotype
gate, 20%; FMO gate, 3,2%) and for the Stro-1 marker (b, isotype gate, 23%; FMO gate, 0,9%). In
polychromatic panels, the only correct gating controls for dimly expressed antigens are FMO controls, even
with perfectly compensated data
3.4.6 Compensation 1. For each experiment, VersaComp (Beckman Coulter) compen-

sation beads are stained with each fluorochrome-labeled anti-
body in individual tubes and incubated for 20 min at room
temperature in the dark following manufacturer’s instructions.
2. Unstained beads are used as an unstained negative control.
Stained beads are washed with BD Lyse Wash Assistant, resus-
pended in staining buffer, kept on ice, and analyzed within 2 h
of processing.
3. For 7-AAD compensation, 1.105 cells are fixed and permeabi-
lized with BD Cytofix/Cytoperm for 20 min, washed with BD
Lyse Wash Assistant, and mixed with 1.105 viable cells to get
two distinct subsets for 7-AADneg and 7-AADpos popula-
tions. Samples are kept on ice and analyzed within 2 h of
processing after a 10 min incubation with 7-AAD.
3.4.7 Cytometer Settings 1. Samples are acquired on a BD FACSCanto II flow cytometer

equipped with 3 lasers (violet [405 nm], blue [488 nm], and
red [633 nm]).
Fig. 4 Examples of FMO/isotype-set gates on density plots. Cells were stained with all antibodies of the panel
except one. Quadrants were established such that the positive events measured represented nonspecific
binding by the fluorochrome-conjugated isotype-matched control. FMO/isotype-set gates are shown here for
CD146 and Stro-1 (a, left) or CD146 and MSCA-1 (a, right). These gates were used to analyze the populations
of the full panel (b)
2. Prior to data acquisition, the photomultiplier tube voltage

(PMT-V) is calibrated to the highest signal to background
ratio by using Cytometer Setup and Tracking (CS&T) beads
and the Cytometer Setup and Tracking software following
manufacturer’s instructions.
Fig. 5 Gating strategy for the exclusion of doublets (a), dead cells (b), and remaining debris and clumps (c)
3. In order to eliminate debris, thresholds are set on FSC channel

between 5000 and 15,000 depending on the sample.
4. The target number of acquired events for each tube of the panel
is 3.105 for in vitro immunophenotyping and 5.104 for in vivo
immunophenotyping. This target is set to 1.105 events for
unstained, isotype FMO and compensation controls.
5. Data are acquired as uncompensated events and recorded as
FCS 3.0 files. Analysis and compensation are performed using
the FlowJo vX software.
3.4.8 Gating Strategy 1. A primary gate is placed on the area versus height signal of the
forward scatter (FSC-A/FSC-H) dot plot to discriminate
between doublets and clumps (Fig. 5 left).
2. A Boolean gate is then set on the 7-AADneg cells, enabling the
analysis of a viable single cell population (Fig. 5 middle).
3. The population is identified by defining the gated population
on a side scatter area signal versus a forward scatter area
(SSC-A/FSC-A) signal dot plot (Fig. 5 right).
4. The number of cells positively stained for a given marker is
determined by the percentage of cells present within a gate
established such that less than 1% of the positive events
measured represent nonspecific binding by the fluorochrome-
conjugated isotype-matched control within FMO control
tubes. Examples of FMO/isotype-set gates on density plots
are shown in Fig. 4.
4 Notes
4.1 Cell Collection 1. Only impacted third molars between Nolla developmental
stages 5 (crown almost completed) and 7 (one third root
completed) were used.
2. Teeth allowed for getting about 20 fragments (explants) from

1 DP.
3. Pooling of outgrowing DP cells from 15 to 20 explants allowed
for the harvest of about 106 cells after 2 weeks of culture.
4. The target number of acquired events for each tube of the
panel was 3.105 for in vitro and 5.104 for in vivo
immunophenotyping.
4.2 Flow Cytometry 1. Sample Processing

Analysis (a) The use of PBS w/o Ca2+ and Mg2+ is mandatory for
optimal efficiency of the TrypLE solution [15].
(b) Lyse Wash Assistant washing can be replaced by classical
washing steps using PBS or staining buffer.
2. Antibody Panel Design and Fluorochrome Selection
(a) Filter selection is a give-and-take process: the use of a
wider bandpass filter can increase the ability to detect a
given fluorochrome but may also increase the amount of
spillover background contributed into that detector from
other neighboring fluorochromes. A good way to visua-
lize these effects is by virtual testing of filter combinations
with a web tool such as the viewer at bdbiosciences.com/
spectra or biolegend.com/spectraanalyzer [16, 17].
(b) The stain indexes supplied by major manufacturers are
calculated for singly reagents, not as part of a cocktail.
As soon as other reagents are added, spectral overlap
becomes an issue.
3. Controls
(a) Isotype controls are antibodies of the same class (isotype)
of immunoglobulins as the specific antibody but are raised
against an antigen which is presumed not to be present on
or in the cells under study. The ideal isotype control
should “match” the specific antibody not only in heavy
chain (IgA, IgG, IgD, IgE, or IgM), subclass, and light
chain (kappa, lambda) class but also in fluorochrome type
and number of fluorochrome molecules per immunoglob-
ulin (F/P ratio). It should have been produced by the
same manufacturing process as the specific conjugate
under investigation and should be presented in the same
formulation (buffer, concentration, preservative, etc.).
References
1. Gronthos S, Mankani M, Brahim J, Robey PG, 2. Huang GTJ, Gronthos S, Shi S (2009) Mesen-
Shi S (2000) Postnatal human dental pulp stem chymal stem cells derived from dental
cells (DPSCs) in vitro and in vivo. Proc Natl tissues vs. those from other sources: their
Acad Sci U S A 97:13625–13630
biology and role in regenerative medicine. J lymphocyte subset immunophenotyping.

Dent Res 88:792–806 Cytometry 42:363–370
3. Kawashima N (2012) Characterisation of den- 10. Baumgarth N, Roederer M (2000) A practical
tal pulp stem cells: A new horizon for tissue approach to multicolor flow cytometry for
regeneration? Arch Oral Biol 57:1439–1458 immunophenotyping. J Immunol Methods
4. Ducret M, Fabre H, Degoul O, Atzeni G, 243:77–97
McGuckin C, Forraz N et al (2016) Immuno- 11. Roederer M (2001) Spectral compensation for
phenotyping reveals the diversity of human flow cytometry: visualization artifacts, limita-
dental pulp mesenchymal stromal cells in vivo tions, and caveats. Cytometry 205:194–205
and their evolution upon in vitro amplification. 12. Hulspas R, O’Gorman MRG, Wood BL, Gra-
Front Physiol 8(7):512 tama JW, Robert Sutherland D (2009) Con-
5. Dominici M, Le Blanc K, Mueller I, Slaper- siderations for the control of background
Cortenbach I, Marini F, Krause D et al fluorescence in clinical flow cytometry. Cyto-
(2006) Minimal criteria for defining multipo- metry B Clin Cytom 76:355–364
tent mesenchymal stromal cells. The Interna- 13. Hulspas R (2010) Titration of fluorochrome-
tional Society for cellular therapy position conjugated antibodies for labeling cell surface
statement. Cytotherapy 8:315–317 markers on live cells. Curr Protoc Cytom
6. Al-Nbaheen M, Vishnubalaji R, Ali D, 54:6.29.1–6.29.9
Bouslimi A, Al-Jassir F, Megges M et al 14. Hughes OR, Stewart R, Dimmick I, Jones EA
(2013) Human stromal (mesenchymal) stem (2009) A critical appraisal of factors affecting
cells from bone marrow, adipose tissue and the accuracy of results obtained when using
skin exhibit differences in molecular phenotype flow cytometry in stem cell investigations:
and differentiation potential. Stem Cell Rev where do you put your gates? Cytometry
9:32–43 75:803–810
7. Lv FJ, Tuan RS, Cheung KM, Leung VY 15. Nestler L, Evege E, Mclaughlin J, Munroe D,
(2014) Concise Review: the surface markers Tan T, Wagner K et al (2004) TrypLE TM
and identity of human mesenchymal stem express: a temperature stable replacement for
cells. Stem Cells 32:1408–1419 animal trypsin in cell dissociation applications.
8. Torre ML, Lucarelli E, Guidi S, Ferrari M, Quest 1:42–47
Alessandri G, De Girolamo L et al (2015) Ex 16. Maecker H, Trotter J (2006) Flow cytometry
vivo expanded mesenchymal stromal cell mini- controls, instrument setup, and the determina-
mal quality requirements for clinical application of positivity. Cytometry 69:1037–1042
tion. Stem Cells Dev 24:677–685 17. Maecker H, Trotter J (2009) Selecting
9. Alamo AL, Melnick SJ (2000) Clinical applica- reagents for multicolor BD flow cytometry.
tion of four and five-color flow cytometry Postepy Biochem 55:461–467
Chapter 9
Dental Pulp Stem Cells: Isolation, Characterization,

Expansion, and Odontoblast Differentiation for Tissue
Engineering
Qing Dong, Yuanyuan Wang, Fatemeh Mohabatpour, Li Zheng,
Silvana Papagerakis, Daniel Chen, and Petros Papagerakis
Abstract
Tissue engineering is an interdisciplinary area offering a promising approach by the use of stem cells
combined with scaffolds and signaling factors for regeneration of damaged or lost tissues. Incorporation
of a sufficient number of cells which do not elicit the immunoreaction in the body is a pivotal element for
successful tissue formation using this method. Stem cells exhibiting strong capacity to self-renew and
differentiate into different cell types are considered as a potent cell source. Among various cell sources,
dental pulp stem cells (DPSCs) are widely under investigation due to the fact that they are simply obtainable
from extracted third molars or orthodontically extracted teeth and show an excellent potential for clinical
application and also their harvesting method is minimally invasive. DPSCs are odontogenic progenitor cells
with clonogenic abilities, rapid proliferation rates, and multiple differentiation potentials. Here, we describe
protocols that allow 1) the isolation of DPSCs from a single tooth; 2) the characterization of human
mesenchymal stem cells markers of DPSCs by flow cytometry; 3) the culture growth of DPSCs in 2D (in cell
culture flasks) and 3D (by 3D printing of cell-laden constructs); and 4) the in vivo evaluation of differentia-
tion potential of DPSCs.
Key words Tissue engineering, Stem cells, Dental pulp, Odontoblast differentiation, Tissue
regeneration
1 Introduction
Tissue engineering is a relatively new multidisciplinary approach

aiming to regenerate tissues which are damaged, diseased, or lost
and restore or ameliorate their function using a combination of
cells, scaffolds, and stimulating factors; thus, it decreases the need
for tissue and organ donation and reduces the risk of transplant
rejection by employing autologous cells from patient’s body
Qing Dong and Yuanyuan Wang contributed equally to this work
91
92 Qing Dong et al.
[1]. Scaffolds which can be fabricated from natural and/or syn-

thetic biomaterials provide a three-dimensional (3D) support for
tissue formation, gradually degrade over time, and are substituted
by newly formed extracellular matrix (ECM) [2]. Numerous meth-
ods exist for scaffold fabrication among which 3D printing techni-
ques produce scaffolds with personalized design and controlled
porosities [3] and also offer the incorporation of living cells and
growth factors into the scaffolds [4].
Signaling factors including bioactive molecules and physical
and mechanical stimulations are able to guide biological activities
such as cell adhesion, proliferation, differentiation, migration, and
ECM production [5, 6]. Incorporation of a sufficient number of
cells which do not evoke the immunoreaction in the body is a
pivotal element for successful tissue formation using this
approach [7].
Stem cells are considered as a potent cell source for engineering
a wide variety of tissues consisting of epithelial surfaces and skeletal
tissues due to their strong capacity to self-renew and differentiate
into different cell types and can be obtained from numerous origins
such as bone marrow, adipose tissue, amniotic fluid, and dental
tissue [8, 9]. There exist several sources of stem cells in oral tissues
which include dental pulp stem cells (DPSC), stem cells from
human exfoliated deciduous teeth (SHED), stem cells from the
apical papilla (SCAP), alveolar bone-derived stem cells (ABSCs),
and dental follicle stem cells (DFSC) among which DPSCs are
widely investigated owing to the fact that they are simply obtainable
and exhibit excellent potential for clinical application and also their
harvesting method is minimally invasive [10]. DPSCs are odonto-
genic progenitor cells and have clonogenic abilities, rapid prolifera-
tion rates, and multiple differentiation potentials (odontogenic,
osteogenic, and adipogenic), providing a suitable cell source for
tissue regeneration [11]. These cells are plastic adherent with a
fibroblast-like morphology and the ability of making clonogenic
colonies when cultured confirms their self-renewal ability
[12]. DPSCs, compared with bone marrow-derived stem cells
(BMSCs) and adipose-derived stem cells (ADSCs), needed longer
time to become fully confluent after isolation, while they exhibited
a significantly higher viability with respect to BMSCs after cryo-
preservation for 14 days and showed a higher level of colony
formation and proliferation rate and mineralization potential
[13, 14]. The major functions of DPSCs are to differentiate into
odontoblasts commencing dentinogenesis, the formation of miner-
alized dentin, and to be involved in enamel formation through
consecutive and mutual interactions with dental epithelial stem
cells [11, 15]; thus, these cells are of great importance in regenera-
tion of enamel, dentin, and pulp tissues. DPSCs can be isolated
from pulp tissues of either deciduous teeth (SHED) or adult teeth
which can be orthodontically extracted teeth or impacted or
Dental Pulp Stem Cells for Tissue Engineering 93
supernumerary ones [16]. Odontogenesis is an important experi-

mental technique for evaluating the odonto-differentiation capacity
of the mesenchymal stem cells derived from oral and maxillofacial
tissue.
Herein, we describe the protocol that allows the systematic
isolation of DPSCs from pulp tissue of a single tooth, either decid-
uous tooth (SHED) or adult teeth which can be orthodontically
extracted tooth, impacted or supernumerary one [16]. Then, the
strategy by which the surface antigens of DPSCs are analyzed by
flow cytometry in order to evaluate the expression of mesenchymal
phenotype in the DPSCs is explained. The biomarkers chosen for
DPSCs analysis usually include CD29, CD34, CD73, CD90,
CD105, and CD146 which are expressed in mesenchymal stem
cells and CD34 and CD45, which are expressed in hematopoietic
stem cells. The expectation is that newly isolated DPSCs are posi-
tive for CD29, CD34, CD73, CD90, CD105, and CD146 and
negative for CD34 and CD45 biomarkers. In addition, steps of
expansion, 2D (in cell culture flasks) and 3D culture (by 3D print-
ing of cell-laden constructs), and odontoblast differentiation of
DPSCs are elaborated. For the in vitro studies, immunohistochem-
ical staining, real-time PCR, and Western blot are usually involved
for odontogenic-specific markers DSP, DPP, and DMP-1 and some
other osteogenic markers ALP, OCN, OPN, and the like. For the
in vivo studies, the mixture of cells and scaffolds is transplanted into
the nude mice for 8 weeks; the pulp-dentin complex like tissue is
recovered and analyzed. Odontogenic differentiation capacity can
be assessed by the results of these two parts together.
2 Materials
2.1 Isolation and 1. Phosphate buffered saline (PBS without Ca++ and Mg++),
Primary Culture of KH2PO4 0.144 g/L, NaCl 9 g/L, Na2HPO4·7H2O
Dental Pulp Stem Cells 0.795 g/L, pH 7.4.
(DPSCs; Fig. 1) 2. Tooth storage medium: Minimum Essential Medium α or PBS,
no calcium, no magnesium with 100 U/mL of penicillin,
100 μg/mL streptomycin, and 0.25 μg/mL amphotericin B.
3. Washing solution: Phosphate buffered saline (PBS) with 2%
antibiotic/antimycotic (200 U/mL of penicillin, 200 μg/mL
streptomycin, and 0.50 μg/mL amphotericin B).
4. Complete growth medium: Alpha Modification of Eagle’s
Medium (α-MEM), supplemented with 15% FBS (fetal bovine
serum), and 1% antibiotic/antimycotic (100 U/mL of penicil-
lin, 100 μg/mL streptomycin) (see Table 1).
94 Qing Dong et al.
Fig. 1 DPSCs have typical fibroblast-like cell morphology (a). Representative colonies are formed after
culturing for 7 days (b)
Table 1
αMEM media components
Media components Final conc in media (mg/L)

CaCl2(anhyd.) 200.00
KCl 400.00
MgSO4(anhyd.) 98.00
NaCl 6800.00
NaHCO3 2200.00
NaH2PO4·H2O 140.00
D-Glucose 1000.00
Lipoic acid 0.20
Phenol red 10.00
Sodium pyruvate 110.00
L-Alanine 25.00
L-Arginine·HCl 127.00
L-Asparagine·H2O 50.00
L-Aspartic acid 30.00
L-Cystine·2HCl 31.00
L-Cysteine HCl·H2O 100.00
L-Glutamic acid 75.00
L-Glutamine 292.00
(continued)
Table 1
(continued)
Media components Final conc in media (mg/L)

Glycine 50.00
L-Histidine HCl·H2O 42.00
L-lsoleucine 52.00
L-Leucine 52.00
L-Lysine·HCl 73.00
L-Methionine 15.00
L-Phenylalanine 32.00
L-Proline 40.00
L-Serine 25.00
L-Threonine 48.00
L-Tryptophan 10.00
L-Tyrosine (disodium salt) 52.00
L-Valine 46.00
L-Ascorbic acid 50.00
Biotin 0.10
D-Ca pantothenate 1.00
Choline chloride 1.00
Folic acid 1.00
L-Inositol 2.00
Niacinamide 1.00
Pyridoxal HCl 1.00
Riboflavin 0.10
Thiamine HCl 1.00
Vitamin B12 1.40
Adenosine 10.00
Cytidine 10.00
Guanosine 10.00
Uridine 10.00
0
2 -Deoxyadenosine 10.00
20 -Deoxycytidine HCl 11.00
0
2 -Deoxyguanosine 10.00
Thymidine 10.00
96 Qing Dong et al.
Fig. 2 Flow cytometry results of mesenchymal stem cell (MSC) markers CD73, CD90, CD105, CD146, CD34,
and CD45. DPSCs were positive for CD29 and CD90 (>90%) and negative for the leukocyte common antigen
CD45 and hematopoietic lineage marker CD34 (<5%)
5. Digestion enzyme solution: 4 mg/mL Dispase II, 3 mg/mL

Collagenase type 1, and 1% antibiotic/antimycotic in PBS.
(a) Cell counting
l 0.4% Trypan blue solution.
(b) Cell passage
l Trypsin-EDTA 0.25%.
l Phosphate buffered solution (PBS) without Ca++ and
Mg++.
l Complete growth medium.
(c) Cell characterization by flow cytometry (Fig. 2)
l Trypsin.
l Phosphate-buffered saline (PBS).
l Fluorescein-conjugated antibodies against the related
markers.
l Mouse Anti-Human CD73.
l Fluorescein-conjugated secondary goat anti-mouse
polyclonal antibody.
l DAPI staining solution.
(d) Odontoblast differentiation
l Odontogenic-differentiation induction media:
10 nmol/L dexamethasone, 10 mmol/L β-glycero-
phosphate, 50 mg/mL ascorbate phosphate,
10 nmol/L 1,25-dihydroxyvitamin D3, and 15% fetal
bovine serum dissolved in Eagle’s medium.
(e) Scaffolds material (Fig. 3)
l Hydroxyapatite ceramic powder.
(f) Sample fix solution
l 4% formalin.
(g) Sample decalcified solution
l Buffered 10% EDTA (pH 8.0).
Fig. 3 A general view of the in vivo experiment was listed (a), the DPSCs cultured with HA. The HA carrier
surfaces were lined with a dentin-like matrix and an interface layer of odontoblast-like cells (b), and new
dentin/pulp-like complex showed positive staining for DSP (c) and VEGF (d)
(h) 3D printing of DPSC-laden alginate hydrogel

l Medium viscosity sodium alginate.
l Polyethylenimine (PEI) (MW: 60000).
l Calcium chloride.
3 Methods
3.1 Isolation Teeth, selected from patients with a healthy medical history, were
erupted third molars or orthodontically extracted permanent pre-
molars with normal dental pulp.
1. Collect the extracted teeth from patients, and store in 50 mL of
tooth storage solution at 4 C for up to 24 h.
2. Aspirate the storage solution, and rinse twice with washing
solution to remove blood and debris.
3. Place tube with the teeth in the ice bucket, and then transfer
one tooth at a time to a 35 mm dish.
4. Excise and discard any gingival tissues using scalpel and
scissors.
5. Split the tooth and pull out the pulp tissue into another sterile
cell culture dish with washing solution.
6. Mince the pulp tissue into 2–4 mm pieces with a surgical
scissors.
98 Qing Dong et al.
7. Transfer the minced pulp tissue into a 15 mL tube, and centri-

fuge tissue at 1200 rpm (G-force 804.96) for 5 min at 4 C,
and discard the supernatant.
8. Add 3 mL enzyme solution into the tube, and incubate for 1 h
in a 37 C shaker.
9. Add 3 mL of complete growth medium to stop the enzyme
digestion procedure.
10. Centrifuge at 1200 rpm (G-force 804.96) for 5 min, discard
supernatant, and resuspend with culture medium 3 mL.
11. Pass the cells through a 70 μm cell strainer and obtain single-
cell suspensions; centrifuge and discard supernatant.
12. Count cells and seed them in complete growth medium.
13. Change medium every 3 days until cells are 70% confluent.
3.2 Cell 1. Subculture DPSCs on 12-chamber slides (2.0 104/well,

Characterization second passage) and grown to confluence 80%.
3.2.1 Immunofluore- 2. Perform standard immunofluorescence methods to detect
scence Staining characterization of stem cells [17].
3. Fix DPSCs initially in 4% paraformaldehyde, and block the
non-specific antigens with PBS containing 10% bovine serum
albumin at room temperature for 1 h.
4. Then incubate the cells with diluted (1:50) primary antibody
CD73, CD90, and CD105 overnight at 4 C.
5. Wash the cells with PBS, and subsequently incubate them with
fluorescein-conjugated secondary goat anti-mouse polyclonal
antibody at room temperature in the dark for 1 h.
6. Perform DAPI staining in the dark for 5 min.
7. Wash the slides with PBS and then analyze using a fluorescence
microscope.
3.2.2 Flow Cytometry 1. Dump old media.

2. Rinse with PBS and dump PBS (see Note 1).
3. Add 0.05% trypsin and incubate at 37 C until cells are lifted
(see Notes 2 and 3).
4. Add DMEM media to the flask (see Notes 4 and 5).
5. Centrifuge solution for 6–7 min at 1300 rpm (G-force 944.71)
(see Note 6).
6. Discard the supernatant.
7. Prepare single-cell suspension (0.5 106/tube), and transfer
them into flow cytometry tube, and incubate them in PBS
containing fluorescein-conjugated antibodies for 60 min.
8. The cells are analyzed using flow cytometry system,
with10,000 events being counted for each case.
3.3 Cell Expansion 1. Dump old media.

(2D Culture) 2. Rinse with PBS and dump PBS (see Note 1).
3. Add 0.05% trypsin and incubate at 37 C until cells are lifted
(see Notes 2 and 3).
4. Add DMEM media to the flask (see Notes 4 and 5).
5. Centrifuge solution for 6–7 min at 1300 rpm (G-force 944.71)
(see Note 6).
6. Dump supernatant and isolate pellet; resuspend pellet in
remaining solution by flicking the vial.
7. Add media based on your desired dilution, and transfer to a
new, labeled, T-75 flask or T-25 flask (see Note 7).
3.4 3D Printing of It has been indicated that alginate scaffolds are able to support the
DPSC-Laden Alginate differentiation of DPSCs after in vivo transplantation [18].
Hydrogel (3D Culture)
1. Prepare a solution of medium viscosity alginate in (3% w/v) in
culture medium under sterile condition (see Note 8).
2. Mix the alginate solution with cells suspended in fresh culture
medium. The final concentration of alginate and density of cells
are 6 106 cells/mL and 2.5% (w/v), respectively.
3. Load the cell-incorporated alginate into the low-temperature
dispensing head (LTDH) of the 3D-BioplotterTM.
4. Print the cells into a cross-linking solution (calcium chloride,
50 mM) into wells of cell culture plates which are treated with
sterile polyethylenimine (PEI) solution in PBS (0.1% w/v) (see
Note 9).
5. After printing hydrogel solution containing cells, CaCl2 solu-
tion is removed, and the construct is washed and cultured in
culture medium in the incubator.
3.5 DPSC Previous studies of solid polymer scaffolds have shown that differ-
Odontogenic entiation of DPSC may be stimulated by the presence of a tooth
Differentiation slice [19].
1. Cut tooth slice which is about 2 mm thick and sterilize the
pulpless tooth slice.
2. Prepare differentiation cell culture media.
3. Centrifuge 1 105 DPSCs, and carefully seed the fabricated
cell pellet in the hole of the tooth slice.
2. Add differentiation cell culture media to DPSCs carefully.
Change the medium every 2 days up to 21 days.
3. Do immunohistochemical staining, real-time PCR, and West-
ern blot to evaluate the odontogenic differentiation.
100 Qing Dong et al.
3.6 In Vivo 1. Centrifuge the DPSCs (1.0 107, third or fourth passage)
Transplantation of after detachment with trypsin to make a cell pellet.
DPSCs 2. Resuspend the pellet into 1 mL of culture medium.
3. Mix the resultant cell suspension with hydroxyapatite
(HA) powder (40 mg), and incubate the cells with HA vehicles
for 3 h at 37 C.
4. Centrifuge the mixture and discard the supernatant to obtain
particles.
5. Implant the vehicles loaded with cells into the dorsal surfaces of
10-week-old mice with severe combined immunodeficiency
(CB-17 scid mice, NIH-bg-nu-xid; Harlan Sprague-Dawley;
Harlan Laboratories, Indianapolis, IN) (see Note 10).
6. Weeks after transplantation, retrieve the implants and fix them
with 4% formalin, decalcified with buffered 10% EDTA
(pH 8.0), and finally embed them in paraffin.
7. Deparaffinize sections (5 mm) and stain them with hematoxy-
lin-eosin.
8. Quantify dentin formation via ImageJ (NIH Image software).
9. The structure of the new formed tissue can be analyzed using
immunohistochemical staining with different markers.
4 Notes
1. The recommended amount of PBS is 5 mL for a T-75 flask and

3 mL for a T-25 flask.
2. The recommended amount of trypsin is 3 mL for a T-75 flask
and 1 mL for a T-25 flask.
3. Check to see if cells are lifted with the microscope or visually by
tapping the side of the flask. You must check on the cells and
tap them every couple of minutes. The cells are susceptible to
death, and their cellular surface antigen expressions are altered
if you leave them in concentrated trypsin for too long.
4. The recommended amount of DMEM media is 5 mL for a
T-75 flask and 3 mL for a T-25 flask.
5. Stir the trypsin/media/cell solution in the flask and transfer to
a 15 mL vial. Label the vial with the appropriate cell line and
any treatments. Remember it is easier if you add the media with
a 10 mL pipette because then you can easily transfer all 10 mL
(or all 6 mL total for a T-25 flask).
6. Remember to balance!
7. Make sure you have a total of 15 mL in the T-75 flask and 5 mL
in the T-25 flask. Label with UMSCC cell line, your name,
current passage number, and the date. Record the dilution you
used in your lab notebook.
8. To make sterile alginate, a low concentration solution of

medium viscosity alginate in deionized water (0.2% w/v) is
prepared and then sterilized via a 0.22 μm filter. The resultant
solution is kept at 40 C freezer overnight and freeze-dried
under sterile condition for 3 days and stored at 4 C.
9. 24 h before printing, the sterile PEI solution (0.1% w/v) is
added into wells (1–2 mL) and kept in the incubator at 37 C.
Before printing, PEI solution is removed and substituted with
cross-linking solution.
10. Sterile operation is extremely important for the step of trans-
plantation. After transplantation, the nude mice should be fed
with glucose 5 mL and high fat and protein food for 1 day.
References
1. Lanza R, Langer R, Vacanti JP (2011) Princi- cells (DPSCs) in vitro and in vivo. Proc Natl
ples of tissue engineering. Academic Press, Acad Sci 97(25):13625–13630
Cambridge, MA 12. Sharpe PT (2016) Dental mesenchymal stem
2. Lee SJ, Yoo JJ, Atala A (2018) Biomaterials and cells. Development 143(13):2273–2280
tissue engineering. In: Clinical regenerative 13. Demirci S, Doğan A, Şahin F (2016) Dental
medicine in urology. Springer, Berlin, pp stem cells vs. other mesenchymal stem cells:
17–51 their pluripotency and role in regenerative
3. Park J et al (2017) Fabrication and characteri- medicine. In: Dental stem cells. Springer, Ber-
zation of 3D-printed bone-like β-tricalcium lin, pp 109–124
phosphate/polycaprolactone scaffolds for den- 14. Nuti N, Corallo C, Chan BMF, Ferrari M,
tal tissue engineering. J Ind Eng Chem Gerami-Naini B (2016) Multipotent differen-
46:175–181 tiation of human dental pulp stem cells: a liter-
4. Ning L, Chen X (2017) A brief review of ature review. Stem Cell Rev Rep 12
extrusion-based tissue scaffold bio-printing. (5):511–523
Biotechnol J 12 15. Tatullo M, Marrelli M, Shakesheff KM, White
5. Lutolf MP, Hubbell JA (2005) Synthetic bio- LJ (2015) Dental pulp stem cells: function, iso-
materials as instructive extracellular microen- lation and applications in regenerative medicine.
vironments for morphogenesis in tissue J Tissue Eng Regen Med 9(11):1205–1216
engineering. Nat Biotechnol 23(1):47 16. Sunil PM, Manikandan R, Muthumurugan
6. Assanah F, Khan Y (2018) Cell responses to TRY, Sivakumar M (2015) Harvesting dental
physical forces, and how they inform the design stem cells-overview. J Pharm Bioallied Sci 7
of tissue-engineered constructs for bone repair: (Suppl 2):S384
a review. J Mater Sci 53(8):5618–5640 17. Bakkar M et al (2017) A simplified and system-
7. Heath CA (2000) Cells for tissue engineering. atic method to isolate, culture, and characterize
Trends Biotechnol 18(1):17–19 multiple types of human dental stem cells from a
8. Bianco P, Robey PG (2001) Stem cells in tissue single tooth. Methods Mol Biol 1553:191–207
engineering. Nature 414(6859):118 18. Cvikl B et al (2017) Response of human dental
9. Atala A, Lanza R, Thomson JA, Nerem R pulp cells to a silver-containing PLGA/TCP-
(2010) Principles of regenerative medicine. nanofabric as a potential antibacterial regener-
Academic Press, Cambridge, MA ative pulp-capping material. BMC Oral Health
10. Chalisserry EP, Nam SY, Park SH, Anil S 17(1):57
(2017) Therapeutic potential of dental stem 19. Sakai VT, Cordeiro MM, Dong Z, Zhang Z,
cells. J Tissue Eng 8:2041731417702531 Zeitlin BD, Nör JE (2011) Tooth slice/scaf-
11. Gronthos S, Mankani M, Brahim J, Robey PG, fold model of dental pulp tissue engineering.
Shi S (2000) Postnatal human dental pulp stem Adv Dent Res 23(3):325–332
Chapter 10
In Vitro Analysis of Intramolecular Signaling Events

in PDLSCs Using Confocal and TIRF Microscopy
Annette Merkel and Anne George
Abstract
The advances in microscopy techniques enable the detection of intracellular molecular processes to be
visualized and analyzed for periodontal ligament stem cells (PDLSCs). Confocal laser scanning microscopy
(CLSM) and total internal reflection fluorescence microscopy (TIRFM) are two well-studied microscopy
techniques that allow an increase in the resolution and contrast of the micrographs and the capability to
pinpoint events at the plasma membrane, respectively. Confocal microscopy achieves its increased resolution
and contrast through a spatial pinhole that hits the plane of the image. TIRF microscopy uses the principle
of incident angles and the refractive index of the substances to study the events occurring at 100–200 nm
range of the cover glass with minimal background interference. Here we describe two methods for
intramolecular signaling visualization upon stimulation with a ligand in normal growth conditions and
mineralization-induced conditions in periodontal ligament stem cells (PDLSCs). These methods are
important for visualizing the signaling events within PDLSCs at a molecular level and thereby understand
the mechanisms by which these cells could be manipulated for tissue engineering and regeneration.
Key words Confocal microscopy, Immunocytochemistry, PDLSCs, TIRFM, Receptor, Ligand
1 Introduction
Periodontal ligament stem cells (PDLSCs) are versatile stem cells

because of their capacity to differentiate into multiple lineages, such
as cementoblasts, fibroblasts, and osteoblasts, making them advan-
tageous for tissue regeneration. Periodontal disease can destroy the
hard tissues of the periodontium, which can lead to bone, peri-
odontal ligament, and eventually tooth loss in those suffering from
the disease [1]. Although PDLSCs possess the ability to differenti-
ate into the various components of the periodontium, their success
in tissue regeneration has been limited due to the lack of funda-
mental knowledge of their molecular processes. With the use of a
ligand, we can visually track the molecular processes of PDLSCs
under normal growth and mineralization conditions with the
sophisticated microscopy techniques of confocal laser scanning
103
104 Annette Merkel and Anne George
Fig. 1 This figure represents an example of determining intracellular trafficking of vesicles in periodontal
ligament stem cells with Rab7, a marker for late endosomes, and glucose-regulated protein-78 (GRP78), a
receptor for dentin matrix protein-1 with confocal microscopy. Periodontal ligament stem cells were serum
starved for 4 h and then treated with rDMP1 at progressing time points. Immunocytochemistry was performed
with antibodies for Rab7 (TRITC) and GRP78 (FITC). The slides were mounted and processed with a Zeiss Laser
Scanning Microscope 510 Meta. Within 10 min, colocalization occurs between GRP78 and Rab7, suggesting
that DMP1 is transported in PDLSCs from early to late endosomes intracellularly
microscopy (CLSM) and total internal reflection fluorescence

microscopy (TIRFM). With these microscopy techniques, the pro-
cesses can be examined on both fixed and live cells.
CLSM is a microscopy technique developed in the late 1980s to
improve both the resolution and contrast of the specimens in
micrographs [2]. CLSM utilizes a spatial pinhole to minimize the
fluorescence emitted throughout the entire specimen in widefield
view and instead focus on a single layer of the cell. Lasers penetrate
the pinhole aperture to hit the cover glass at that specific focus
point size of the pinhole (usually between 0.25 and 0.8 μm)
[3]. This small focus point decreases the background fluorescence
emission, thus producing a clear and crisp image. The advantages
that CLSM provides have made it a popular microscopy technique
in the life sciences field since its inception. Intracellular interactions
can be visualized with CLSM by colocalization methods (see Fig. 1).
CLSM provides clear images that can pinpoint the intracellular
localization and molecular interactions within PDLSCs.
CLSM allows the user to visualize the fluorescence of one
optical section of the specimen, and thus one entire focal plane is
in view; however, the TIRFM technique utilizes the total internal
reflection angle to visualize events occurring 100–200 nm from the
cover glass [4, 5]. In TIRFM, the cells can be live or fixed to a glass
bottom dish with media or 1 PBS in the dish. A large refractive
index between the two interfaces is vital to achieve a critical angle
that causes a total internal reflection within the glass slide [6]. The
specificity of the depth that is achieved with the TIRF angle allows
events to be seen at the plasma membrane, such as receptor-ligand
interactions, endocytosis, and exocytosis [4]. This sophisticated
technique greatly reduces the background fluorescence, and the
limited depth decreases the signal-to-noise ratio producing crisp
images of the events at the plasma membrane (see Fig. 2). TIRFM is
advantageous for studying biomineralization events such as
PDLSCs and Microscopy Techniques 105
Fig. 2 This is an example of TIRF images of PDLSCs. The PDLSCs were transfected with pCDH-GRP78-GFP for
2 days prior to treatment with dentin matrix protein-1 (DMP1) at various time points. The PDLSCs were fixed
and washed in PBS before TIRF processing. These images represent an increase of GRP78 to the plasma
membrane of PDLSCs with DMP1. These images gives insight to what happens at the membrane with DMP1
during biomineralization, and it can be used with other cell types as well
exocytosis and endocytosis in PDLSCs because it provides evidence

for the receptor-ligand interactions at the plasma membrane.
In this chapter, we present two different microscopy techniques
to study ligand-receptor interactions and cellular trafficking in
PDLSCs. These techniques can provide insight into the molecular
processes of PDLSCs upon stimulation with a ligand or during
mineralization. The methods described will provide a framework
for understanding these molecular processes and thereby lead to a
better understanding of how to utilize these cells in biomineraliza-
tion and tissue regeneration.
2 Materials
2.1 Cell Culture 1. Periodontal ligament stem cell (PDLSC) media: α-MEM
media, 500 mM glutamine, 100 penicillin/streptomycin
stock, fetal bovine serum. Add 75 mL of FBS, 5 mL of gluta-
mine, and 5 mL of penicillin/streptomycin to 500 mL of α-
MEM media. Filter media prior to use.
2. Ascorbate phosphate: Prepare 10 mg/mL of ascorbate phos-
phate in water to make a 100 stock. The final working con-
centration will be 100 μg/mL (see Note 1).
3. β-Glycerophosphate: Prepare 1 M of β-glycerophosphate in

H2O to make a 100 stock. The final working concentration
will be 10 mM.
4. Dexamethasone: Prepare 1 μM in 1 PBS. The final working
concentration will be 10 nM.
5. PDLSC mineralization media: Prepare the media as described
previously. Add the ascorbate phosphate, β-glycerophosphate,
and dexamethasone to their appropriate final working concen-
trations from the 100 stocks. Filter media.
6. PDLSC serum-free media: α-MEM media, 500 m glutamine,
100 penicillin/streptomycin. Add 5 mL of glutamine and
5 mL of penicillin/streptomycin to 500 mL of α-MEM
media. Filter media prior to use.
7. 1 PBS.
8. 6-well cell culture plates.
9. 25 mm cover glass.
10. 35 mm glass bottom dish tissue culture treated.
11. Transfection reagent of choice (see Note 2).
12. OPTI-FREE Media.
13. 0.5% trypsin-EDTA.
14. Ligand of interest.
15. Plasmid with fluorescent marker.
16. Incubator: 37 C and 5% CO2.
17. Automated cell counter or hemocytometer.
18. Trypan blue.
2.2 Immunocyto- 1. 10% Formalin.

chemistry 2. 0.5% Triton X-100 in 1 PBS.
3. 5% BSA in 1 PBS.
4. 1 PBS.
5. VectaShield Mounting Media for fluorescence with DAPI.
6. Glass slides.
7. Antibodies of interest.
8. Appropriate secondary antibodies.
9. Nail polish.
2.3 Processing 1. Confocal laser scanning microscope (CLSM).

2. Total internal reflection fluorescence microscopy (TIRFM).
3. 37 C incubator adaptor for TIRF microscope.
3 Methods
3.1 Methods 1. Plate approximately 1.0 106 cells on 10 mm tissue culture

for CLSM plates in PDLSC media, and incubate (see Note 3).
3.1.1 Cell Culture 2. Observe the PDLSCs under a light microscope to determine if
the cells are confluent to be passaged (approximately 80–90%
confluency). The PDLSCs reach this stage of confluency within
2–3 days of growth with media change every other day.
3. Prepare two 6-well plates with a cover glass in each well. Place
plates under UV light for 30 min.
4. Once the cells have reached confluency, remove the PDLSC
media, and add 2.5 mL of trypsin-EDTA to the plate. Incubate
the cells for 4–5 min until the cells start lifting off the plate.
5. Add 7.5 mL of PDLSC media to stop the trypsin reaction, and
transfer to a 15 mL conical tube.
6. Take approximately 10 μL of the trypsin-cell media to use for
cell counting (see Note 4).
7. Centrifugate the cells at 300–1000 g for 5 min to collect the
cell pellet.
8. Remove the supernatant and resuspend the cell pellet in
PDLSC media.
9. Seed 50,000 cells in each 6-well plate with 2 mL of media in
each well.
10. Incubate the plates at 37 C.
11. Once a plate has reached 60% confluency, remove the PDLSC
media from one plate, and add 2 mL of PDLSC mineralization
media to each well. Let the cells continue growing for 2 days.
3.1.2 Treatment 1. Once the cells have reached 80% confluency, remove the
PDLSC media from one plate and the PDLSC mineralization
media from the other, and wash with 2 mL per well of 1 PBS
twice (see Note 5).
2. Add 2 mL of serum-free PDLSC media to each well, and
incubate for 4 h (see Note 6).
3. Add 250 ng of the ligand to be investigated to each well at
60, 30, 15, 10, and 5 min. One well will be the control well (see
Note 7).
4. Remove the media and wash twice with 1 PBS.
5. Fix the cells with 10% formalin for at least 2 h (see Note 8).
3.1.3 Immunocyto- 1. Remove the 10% formalin, and wash the cells with 1 PBS four
chemistry times.
2. Permeabilize the cells with 0.5% Triton X-100 in PBS for

30 min RT.
3. Wash four times with 1 PBS.
4. Block the cells with 5% BSA for an hour at RT (see Note 9).
6. Make 1 mL of primary antibody in 5% BSA for each well with
the proper dilution of the antibody.
7. Apply the primary antibody with the dilution recommended.
Leave the primary antibody on overnight at 4 C.
9. Make 1 mL of fluorescent secondary antibodies in 5% BSA
for each well with the proper dilution of the antibody (see
Note 10).
10. Apply the secondary antibody with the dilution recommended
for 1 h at room RT.
12. Mount the cover glass on glass slides for imaging. Use one
drop of VectaShield Mounting Media for fluorescence with
DAPI prior to mounting on the coverslips.
13. Seal the glass coverslips with nail polish and allow the slides to
dry. Store at 20 C until further processing.
14. Visualize and image the slides with CLSM.
3.2 Methods for 1. Repeat the cell culture steps 1–8.

TIRFM 2. Seed 50,000 cells in 6–35 mm dishes treated for tissue culture.
3.2.1 Cell Culture Allow growth to 80% confluency (approximately 2–3 days).
3.2.2 Transfection 1. Warm the transfection reagent to room temperature prior to

use (see Note 11).
2. Mix plasmid of interest containing GFP with the transfection
reagent in approximately 50 μL of OPTI-FREE media. Let the
plasmid-transfection mixture sit at RT for 20 min (see Notes 12
and 13).
3. Pipette the plasmid-transfection mixture into each 35 mm dish.
4. Change the PDLSC media the following morning.
5. With a fluorescent microscopy, check the transfection efficiency
after 2 days (see Note 14).
3.2.3 Treatment for Fixed 1. Remove the PDLSC media and replace with serum-free
TIRF Samples PDLSC media for at least 4 h.
2. Add 250 ng of the ligand to each well for 0, 5, 10, 15, 30, and
60 min. One well will be the control well (see Note 7).
3. Remove the media and wash with 1 PBS twice.

4. Fix the cells with 10% formalin for at least 2 h (see Note 8).
3.2.4 Treatment for Live 1. Remove the PDLSC media, and replace with PDLSC serum-
TIRF Samples free media for at least 4 h (see Note 6).
2. Place one plate into the 35 mm dish holder on the microscope
in a 37 C chamber.
3. Treat the plate with the ligand of interest, and capture the
TIRF images at the time points of 0, 5, 10, 15, 30, and
60 min (see Note 15).
3.2.5 Processing with 1. Wash each 35 mm dish with 1 PBS four times. Leave the
TIRFM for Fixed Samples remaining wash as 1 PBS prior to processing.
4 Notes
1. L-Ascorbic acid can also be used instead of ascorbate phos-

phate, but it should be made fresh prior to use. Ascorbate
phosphate is more stable and can be stored in aliquots at
20 C for an extended period of time.
2. For increased transfection efficiency, try using DNAfectin as
the transfection reagent over Lipofectamine for stem cells,
especially PDLSCs and DPSCs.
3. All PDLSC incubations are to be at 37 C with 5% CO2 unless
otherwise stated.
4. Cell counting can be done with an automated cell counter.
There should be a 1:1 ratio between media and trypan blue.
A hemocytometer can also be used.
5. All washing steps should be done at room temperature with
light shaking for 15 min.
6. Prior to the treatment with the ligand, the cells should be in
serum-free (no FBS) alpha-MEM period for at least 4 h. This
starvation period can be extended overnight if necessary.
7. The amount of ligand treatment is dependent upon the ligand
itself. For optimal conditions, use 250 ng as a starting point,
and adjust the concentrations accordingly.
8. The fixation period should be at least 2 h at room temperature.
The optimal results with PDLSCs have been with overnight
fixation at 4 C.
9. The blocking step can be anywhere between 1 and 5 h at RT or
4 C overnight.
10. The immunocytochemistry— step 9—should be done in a

dark room and washing steps performed in foil-covered con-
tainers due to the light-sensitive fluorescent antibodies.
11. Warm the transfection reagent to RT for approximately
20 min.
12. The plasmid-transfection reagent mixture should be at RT for
at least 20 min, but it can be at RT for up to 3 h.
13. The transfection reagent to plasmid ratio will require optimi-
zation. Our results show the most optimal transfection effi-
ciencies have been achieved with 4 μL of 1 ng DNA plasmid
with 1 μL of DNAfectin in 50 μL.
14. The most optimal transfection efficiencies were obtained after
2 days in PDLSCs.
15. The ligand can be pipetted into the dish while it is incubating.
Another strategy could be to have it injected into the dish while
incubating on the microscope. This is dependent upon the
microscope equipment available.
References
1. Zhu W, Liang M (2015) Periodontal ligament 4. Fish KN (2009) Total internal reflection fluores-
stem cells: current status, concerns, and future cence (TIRF) microscopy. Curr Protoc Cytom
prospects. Stem Cells Int 2015:1–11. https:// 12(12.18). https://doi.org/10.1002/
doi.org/10.1155/2015/972313 0471142956.cy1218s50
2. White JG, Amos WB (1987) Confocal micros- 5. Paddock SW (1999) Confocal laser scanning
copy comes of age. Nature 328(6126):183–184. microscopy. BioTechniques 27(5):992-6–998-
https://doi.org/10.1038/328183a0 1002, 1004
3. Prasad V, Semwogerere D, Weeks ER (2007) 6. Mattheyses AL, Simon SM, Rappoport JZ
Confocal microscopy of colloids. J Phys Con- (2010) Imaging with total internal reflection
dens Matter 19:113102–113125. https://doi. fluorescence microscopy for the cell biologist. J
org/10.1088/0953-8984/19/11/113102 Cell Sci 123(Pt 21):3621–3628. https://doi.
org/10.1242/jcs.056218
Chapter 11
A Mouse Model to Study Reparative Dentinogenesis

R. C. Babb, D. Chandrasekaran, L. K. Zaugg, and P. T. Sharpe
Abstract
Different animal models have been introduced recently to study the process of reparative dentinogenesis in
response to injury-induced pulp exposure. Using a mouse model is advantageous over other animal models
since mice can be genetically manipulated to examine specific cellular pathways and lineage trace the
progeny of a single cell. However, enabling a standardized molar damage in mice is demanding due to
the small size of the teeth compared to the available dental instruments. Here we describe a reproducible
and reliable in vivo model that allows us to study dentinogenesis in the first maxillary mouse molar.
Key words Dentinogenesis, Molar damage, Tooth repair, Mouse model, In vivo
1 Introduction
Reparative dentinogenesis in response to injury-induced pulp expo-

sure involves the recruitment of stem/progenitor cells to the site of
damage. These stem/progenitor cells differentiate into new
odontoblast-like cells that deposit reparative dentin and hence
form a protective seal preventing the pulp from exposure to the
external environment [1, 2]. Several in vivo animal models have
been used to study dentinogenesis, including the rat, dog, monkey,
and ferret [3–6]. These in vivo models are superior to in vitro and
ex vitro tooth models of repair, as the specific 3D spatial arrange-
ment of cells in the dentin-pulp microenvironment remains intact.
The dentin-pulp microenvironment releases regenerating signals,
which mediate the recruitment, differentiation, and proliferation of
stem/progenitor cells into reparative odontoblasts [7].
The mouse is not routinely used to investigate reparative den-
tinogenesis, due to the technical difficulty of creating damage with
standard dentistry equipment and reagents, owing to the small size
of the teeth. Using a mouse model to study dentinogenesis is
advantageous over other animal models since mice can be geneti-
cally manipulated to examine specific cellular pathways and lineage
trace the progeny of a single cell.
111
112 R. C. Babb et al.
We have developed a reliable in vivo model that allows us to

study dentinogenesis in the mouse molar. A cavity is created in the
center of the first maxillary molar, which is capped with mineral
trioxide aggregate (MTA) and glass ionomer (GI) to promote
dentinogenesis [8, 9]. MTA is used because it is nontoxic to cells,
has antibacterial activity, and stimulates the release of bioactive
dentin matrix proteins. However, MTA is highly soluble and
requires a prolonged setting time. To overcome this, a small layer
of a fast-setting GI is placed over the MTA and covered with a
protective sealant using a dental adhesive to form a long-term
bacterial resistant seal.
2 Materials
2.1 Maxillae Molar Prepare all solutions using ultrapure water and analytical grade
Cavity Preparation reagents. Store all reagents at room temperature (unless indicated
Components otherwise).
1. Hypnorm (fentanyl/fluanisone—VetaPharma Ltd., Leeds, UK).
2. Hypnovel (midazolam hydrochloride, Roche).
3. Carbide bur (FG ¼) (Henry Schein, Kent, UK).
4. Mouth retractor with slings, length 7 cm (InterFocus Ltd.,
Linton, UK).
5. Phosphate-buffered saline (PBS).
6. Mineral trioxide aggregate (MTA) (ProRoot MTA, Dentsply,
Tulsa, Oklahoma, USA).
7. Glass ionomer cement (GI) (Ketac™ Cem, 3 M ESPE, Seefeld,
Germany).
8. Dental adhesive (SU) (Scotchbond™ Universal Adhesive,
3 M ESPE).
9. Microbrush (Microbrush, Grafton, WI, USA).
10. Buprenorphine, Vetergesic® Multidose (Centaur Services,
Somerset, UK).
2.2 Decalcification of 1. Four percent paraformaldehyde (PFA): Dissolve 4 g of PFA in

Maxillae Components 100 mL of PBS.
2. Nineteen percent ethylenediaminetetraacetic acid (EDTA):
Dissolve 100 g of EDTA tetrasodium salt dihydrate and
90 g EDTA disodium salt dihydrate in 900 mL distilled
water, pH to 7.4, with hydrochloric acid, and adjust the
volume to 1 L.
Reparative Dentinogenesis 113
2.3 Process Maxillae 1. Methylated spirits (IMS).

to Wax Components 2. Ultraplast wax (Solmedia, Shrewsbury, UK).
3. TruBond 380 (VWR International Ltd., Leicestershire, UK).
2.4 Masson’s 1. Histo-Clear II (AGTC Bioproducts t/a National Diagnostics,

Trichrome Staining Hessle, UK).
Components 2. Weigert’s hematoxylin (Solmedia).
3. Biebrich scarlet-acid fuchsin solution: Dissolve 1 g of acid
fuchsin (MP Biomedicals, Cambridge, UK) in 100 mL of
Milli-Pure water. Dissolve 10 g Biebrich scarlet in 1 L Milli-
Pure water. To make 100 mL Biebrich scarlet-acid fuchsin
solution, combine 89 mL of 1% Biebrich scarlet with 10 mL
1% acid fuchsin and 1 mL glacial acetic acid.
4. Phosphomolybdic-phosphotungstic acid solution: Dissolve
3.125 g phosphomolybdic and phosphotungstic acids in
250 mL Milli-Pure water.
5. Aniline blue solution: Dissolve 2.5 g aniline blue in 99 mL of
Milli-Pure water and 1 mL of glacial acetic acid.
6. Neo-Clear (Merck Chemicals Ltd., Nottingham, UK).
7. Neo-Mount (Merck Chemicals).
3 Methods
3.1 Maxillae Molar Under the Animal (Scientific Procedures) Act 1986, anyone wish-
Cavity Preparation ing to carry out a regulated procedure on a protected animal must
hold an establishment, project, and personal license. Use sterilized
surgical instruments, solutions, and cotton pellets only for the
following procedures.
1. Anesthetize mice using a combination of Hypnorm, sterile
water, and Hypnovel in the ration 1:2:1. Inject solution at
5 mL/kg intraperitoneally (IP) (see Note 1).
2. Position the anesthetized mouse on the microscope stage.
Open the oral cavity with a mouth retractor; first use a tweezer
to place the tongue over the mandible incisor, and then secure
the slings of the retractor between the mandible and maxilla
incisors (Fig. 1a). Position blunt-ended curved tweezer in
either side of the maxillary first molars to prevent the cheeks
from obscuring the view and from any damage during the
procedure (Fig. 1b, m; see Note 2).
3. Clean the maxillary first molars with a cotton plug soaked in
PBS. Create a cavity in the center of the maxillary first molars
Fig. 1 Images (a–d) present the general sequence of the molar drill damage procedure, images (e–l) show
detailed intraoral views, and images (m–p) represent the instruments and materials used. Mice are positioned
using a carbide bur (FG ¼) coupled to a dental high-speed

handpiece (Kavo Super Torque LUX 2 640B) (Fig. 1c, f). Stop
drilling when the pulp is visible under the dentin roof (Fig. 1g).
Use a 27G 3/4 needle to expose the pulp chamber (Fig. 1d, h, i;
see Notes 3–6).
4. Prepare the mineral trioxide aggregate (MTA) by mixing a
small amount of MTA powder with PBS to get a consistency
of wet sand (Fig. 1n). Use the end of a 27G 3/4 needle to
collect some MTA, and place it on the exposed pulp (Fig. 1j, o).
Next, prepare the glass ionomer (GI) by mixing a small amount
of the liquid provided in the kit to get a toothpaste-like consis-
tency (see Note 7). Collect a small amount of GI with the end
of a 27G 3/4 needle, and apply it to the molar to seal the cavity
(Fig. 1k).
5. Squeeze one drop of the dental adhesive (SU) on a clean
surface, and cover it from light until use. Massage SU gently
for 20 s over set GI using a microbrush (Fig. 1p, l). Dry with
slight air pressure for 5 s and light cure for 10 s with any dental
light-curing device.
6. Inject buprenorphine (0.05–0.1 mg/kg, IP) for pain relief, and
allow mice to recover in a warm environment.
3.2 Decalcification of 1. Sacrifice mice and dissect out the maxillae.

Maxillae 2. Fix the maxillae in 4% paraformaldehyde (see Note 8) for 24 h
at 4 C.
3. Wash the fixed maxillae with PBS three times for 5 min. At this
stage the maxillae can be imaged by micro-CT (Fig. 2).
4. Demineralize the maxillae in 19% EDTA (see Note 9).
3.3 Process Maxillae 1. Wash decalcified maxillae with PBS three times for 5 min, and
to Wax go through a series of 30, 50, and 70% ethanol; store at 4 C
O/N for each concentration.
2. Dehydrate maxillae through a graded methylated spirits (IMS)
series, 70, 90, and 100% methylated spirits, four changes each
for 2 h. Clear in xylene (three changes, 2 h each), and infuse
Fig. 1 (continued) under the microscope, and the mouth is opened using a mouth retractor and tweezers (a, b,
and m) to expose the maxillary first molars (e, first molars indicated by asterisks). A cavity is made in the
center of the maxillary first molar using a carbide bur (c, f, and g), and the pulp chamber is exposed using a
27G 3/4 needle (d, h). The damage measures approximately 0.1 mm (i, scale (dental periodontal probe)
indicates 1 mm for metal plus black line). The cavity is capped with MTA (j) and sealed with glass ionomer
cement (k). The MTA is mixed with water to a consistency of wet sand (n) and a needle is used to apply the
material on the exposed pulp (o). Finally, a dental adhesive is used to protect the restoration (l and p). Pictures
were taken using a Canon EOS 60D and a macro objective (Canon EF-S 60 mm f/2.8 Macro USM) with macro
ring flash (Canon MR-14EX)
Fig. 2 Micro-computed tomography imaging (μCT) of injured maxillary first molar. Sagittal section (a), frontal
section (b), and transverse section (c) of damaged maxillary molar. The cavity (~100 μm) is located at the
center of the occlusal surface; asterisks (a) indicate the capping material MTA, GI indicates the glass ionomer
restoration; scale bar (c) represents 1 mm
with Ultraplast wax at 60 C (three changes, 2 h each) in a

Leica ASP300 tissue processor.
3. Embed the infused maxillae into wax blocks.
4. Cut 8-μm-thick sections using a microtome (Leica RM2245,
blade angle 5 ) (see Note 10).
5. Transfer the sections onto glass slides.
3.4 Masson’s 1. Dewax sections in Histo-Clear II (2 10 min), and rehydrate

Trichrome Staining through decreasing concentrations of IMS, 100, 90, 70, and
50%, for 2 min each.
2. Rinse the sections in water briefly, stain with Weigert’s hema-
toxylin for 10 min, and rinse in running water for 10 min.
3. Stain the sections with Biebrich scarlet-acid fuchsin solution for
4. Differentiate and mordant sections in a mix of 5% phosphomo-
lybdic acid and 5% phosphotungstic acid until the collagen no
longer stains red (about 15 min) (see Note 12).
5. Stain with aniline blue solution for 10 min (see Note 13).
6. Differentiate in 1% acetic acid for 2 min.
7. Dehydrate through 90 and 100% IMS for 5 min each, followed
by three changes in Neo-Clear, and permanently mount in
Neo-Mount (Fig. 3).
Fig. 3 Reparative dentinogenesis of a damaged maxillary first molar 6 weeks after injury visualized by Masson
trichrome staining: collagen (blue), cytoplasm (red), and nuclei (black); asterisk represents the site of damage.
Reparative dentinogenesis can be seen at the site of damage. The newly formed dentin stained darker blue
than the preexisting secondary dentin; furthermore, the dentin tubules are sparse and irregular in pattern, and
some cellular inclusions can be seen, consistent with the appearance of tertiary dentin. Newly formed dentin
can also be seen under the molar cusps, perhaps as a reaction to the drilling procedure. The pulp is vital, and
there are no signs of inflammation. These observations demonstrate that the mouse maxillary first molars can
be used to study the formation of new dentin in response to damage
4 Notes
1. All anesthetized animals must be constantly attended and mon-

itored to assess the adequate level of anesthesia.
2. Be careful to position the microscope light away from the
mouse’s eyes to prevent damage and possible blindness.
3. Do not continuously drill through the central cusp; instead, use
short drill intervals of 2–3 s. Pay special attention not to touch
any soft tissue (cheeks, tongue, etc.) with the bur shaft to avoid
burn damage. In between drill intervals, apply one to two drops
of PBS to prevent overheating and burning of the enamel and
dentin. Dry the PBS using a sterile cotton plug before drilling
again.
4. Identifying the dentin roof is difficult and comes with practice.
In our experience, the dentin roof has been reached when the
dentin begins to appear translucent, almost glasslike, and the
yellow/pink pulp can be seen underneath. At this point, gently
touch the needle tip against the dentin roof, and a hole should
easily be created that just breaks through the dentin (Fig. 1i). If
you touch the needle on the dentin and no hole appears, do not
push hard; instead, drill again and retry until a hole can be
created. If the needle is pushed too deep and goes through the
pulp chamber floor, repair is unlikely to occur.
5. The angle of the bur will be different when drilling the left and
the right maxillary molar to prevent the drill head obscuring
the view; in our experience, drilling the left molar is more
difficult than the right.
6. Occasionally apply a drop of PBS to the mouse’s tongue to
prevent it from drying and sticking to the mouth retractor.
7. When the MTA is applied, compress it using a sterile cotton
bud to ensure that it is in contact with the exposed pulp. Allow
the MTA to dry for a few minutes so that it has a sand-like
appearance before applying the GI. When applying the GI,
make sure that it is in contact with the MTA and not with the
exposed pulp as it is toxic. Allow the GI to fully dry (about
3–4 min), and form a hard seal prior to applying the dental
adhesive (SU). The capping seal should stay in place for at least
2 weeks.
8. Measure 100 mL PBS into a conical flask containing 4 g of
paraformaldehyde. Cover tightly with parafilm, and place the
flask on top of the hotplate/stirrer inside. Stir until dissolved.
This should all be performed in the fume hood. PFA can be
stored at 20 C.
9. Decalcification should be carried out at 4 C with agitation.
Replace the EDTA solution every 3 days. Stop decalcification
when maxillae are soft enough to be slightly bent with twee-
zers; this can take between 2 and 4 weeks.
10. Chill the wax blocks on ice before sectioning. We orientate the
wax block to obtain sagittal sections. If the sections start to
tear, wipe wax block with a water-moistened tissue between
each section.
11. Prepare Biebrich scarlet-acid fuchsin solution just before use.
12. Discard phosphomolybdic-phosphotungstic acid solution
after use.
13. Aniline blue solution can be used up to five times before
discarding.
Acknowledgments
Research was supported by the Swiss National Science Foundation.
References
1. Lesot H, Smith AJ, Tzafias D, Bègue-Kirn C, 2. Goldberg M, Smith AJ (2004) Cells and extra-
Cssidy N, Ruch JV (1994) Biological active mol- cellular matrices of dentin and pulp: a biological
ecule and dental tissue repair, a comparative basis for repair and tissue engineering. Crit Rev
review of reactionary and reparative dentinogen- Oral Biol Med 15:13–27
esis with induction of odontoblast differentia- 3. Costa CA, Oliveira MF, Giro EM, Hebling J
tion in vitro. Cell Mater 4:119–218 (2003) Biocompatibility of resin-based materials
used as pulp-capping agents. Int Endod J dentin matrix components. Arch Oral Biol
36:831–839 39:13–22
4. Holland R, de Souza V, Nery MJ, Otoboni Filho 7. Yu T, Volponi AA, Babb R, An Z, Sharpe TP
JA, Bernabe PF, Dezan E Jr (1999) Reaction of (2015) Stem cells in tooth development, repair
dogs’ teeth to root canal filling with mineral and regeneration. Curr Top Dev Biol
trioxide aggregate or a glass ionomer sealer. J 30:213–222
Endod 25:728–730 8. Hilton TJ (2009) Keys to clinical success with
5. Tarim B, Hafez A, Cox C (1998) Pulpal pulp capping: a review of the literature. Oper
response to a resin-modified glass-ionomer Dent 34:615–625
material on non-exposed and exposed monkey 9. Okiji T, Yoshiba K (2009) Reparative dentino-
pulps. Quintessence Int 29:535–542 genesis induced by mineral trioxide aggregate: a
6. Smith AJ, Robias RS, Cassidy N, Plant G, review from the biological and physicochemical
Browne RM, Begue-Kirn C, Ruch JV, Lesot H points of view. Int J Dent 2009:12. https://doi.
(1994) Odontoblast stimulation in ferrets by org/10.1155/2009/464280
Chapter 12
Multiwalled Carbon Nanotubes for Dental Applications

Petros Kechagioglou, Eleftherios Andriotis, Petros Papagerakis,
and Silvana Papagerakis
Abstract
Multiwalled carbon nanotubes (MWCNTs) are a particularly promising drug delivery system due to their
high surface area allowing high-protein loading, their stability under biological conditions, and their unique
interaction with cellular membranes. Studies have shown that covalent attachment of polyethylene glycol
(PEG) improves biocompatibility and enhances surface hydrophilicity properties, suggesting that PEGy-
lated MWCNTs are efficient and toxic-safe drug delivery systems. So far, CNTs are used for a broad range of
applications in dentistry, especially for dental tissue repair and restorative. Here we present a protocol of
protein immobilization onto MWCNTs and describe the procedure for delivering them into the cells after
characterization of the nanotubes.
Key words Carbon nanotubes, Protein delivery system, Dental applications
1 Introduction
Several delivery systems are developed with limited success due to

low-protein loading, size control, and toxicity. Multiwalled carbon
nanotubes (MWCNTs) are used in biological systems due to their
ability to easily penetrate through the cell membrane, its sustained
capacity, and distribution within cells. Additional studies have
shown that pristine MWCNTs exhibit increased antimicrobial
activity due to their structure against a variety of pathogens
[1–4]. They promote mechanical disruption of the microbial cell
wall by “piercing” or “wrapping” of pathogens and induce intra-
cellular oxidative stress [1].
There are several types of carbon nanostructures, such as single-
and multiwalled carbon nanotubes. Multiwalled carbon nanotubes
(MWCNTs) have a diameter up to 20 nm [5] and are well studied
for a variety of applications, including protein, gene, and drug
delivery as well as diagnostics [6–9]. MWCNTs have emerged as
promising candidates because of their several advantages over other
nano-sized delivery systems, such as an exceptionally high drug
121
122 Petros Kechagioglou et al.
loading capacity due to their high surface area and their ability to
incorporate additional therapeutic and diagnostic molecules, either
on the surface or their inner core. Moreover, they have a unique
ability to interact with cellular membranes. Specifically, some types
of CNTs have been reported to enter mammalian cells by an
endocytosis-independent, “needlelike” penetration mechanism,
which allows for direct cytoplasmic delivery of conjugated mole-
cules [10, 11], while endocytosis is the commonly suggested mech-
anism for MWCNTs [12, 13]. However, there are various
internalization mechanisms of CNTs that are not well known yet
as their cellular uptake depends on several properties.
Despite the advantages of CNTs, they usually must be functio-
nalized for biological applications, in order to render them soluble
in water, improve their biocompatibility, and conjugate biological
and bioactive molecules. A widely used type of such modification is
based on the use of surfactants (Tween-20, Triton X-100, sodium
dodecyl sulfate (SDS)), which attach to or wrap around the gra-
phitic sidewalls of carbon nanotubes by non-covalent interactions
[14]. The most successful strategy employs polyethylene glycol
(PEG) as modifying polymer. PEG offers several properties such
as high dispersibility and stability and prolonged blood circulation
time and prevents opsonization which led to important results in
oncology, neurology, vaccination, and imaging [15].
Meanwhile, various studies investigated the applications of
CNTs in dentistry. Hahn et al. indicated that CNTs had the ability
to improve the mechanical and biological performances of hydroxy-
apatite coatings which is the main substance of our teeth [16]. The
increased stiffness of hydroxyapatite/MWCNT composite was also
revealed by Khan et al. [17]. Another application of CNTs in
dentistry is in bone engineering. A representative study has shown
that CNTs can enhance the mechanical properties of polymethyl
methacrylate (PMMA), a material for bone cement and dental
prostheses [18]. Moreover, Tsukasa et al. in 2009 revealed the
selectivity of CNTs to attach the surfaces of dentin and cementum
but not to enamel, due to the exposed collagen fibers at these two
surfaces [19].
2 Materials
Oxidized multiwalled carbon nanotubes (ox-MWCNTs), D L

9.5 1.5 nm.
O,O0 -Bis(3-aminopropyl) polyethylene glycol (1500 g/Mol),
(C2H4O)nC6H16N2O.
N-Hydroxysuccinimide (NHS) C4H5NO3.
N-(3-Dimethylaminopropyl)-N0 -ethylcarbodiimide hydrochloride
(EDC), C8H17N3·HCl.
Multiwalled Carbin Nanotube 123
Phosphate-buffered saline (PBS) 1, pH 7.2.

0.45 μm polycarbonate filters.
3 Methods
3.1 Functionalization 1. 5 mg oxidized MWCNTs are sonicated for 10 min in 1 mM

of MWCNTs with Bis sterilized phosphate-buffered saline (PBS) pH 7.2.
(3-Aminopropyl) 2. To this suspension is added equal volumes (10 mL) of 2 mM
Polyethylene Glycol EDC/5 mM NHS and incubated for 30 min at room temper-
(Fig. 1) ature while stirring (see Note 1).
Fig. 1 Modification reactions of multiwalled carbon nanotubes (MWCNTs) and conjugated protein
3. 3 mL of 6 mM bis(3-aminopropyl) PEG is added, and the

solution is sonicated for 3 h at room temperature (see Note 2).
4. The PEGylated ox-MWCNTs are filtered twice using a
0.45 μm polycarbonate filters.
5. Wash abundantly with ultrapure water to remove excess
reagents.
6. Lyophilized to get purified PEG-MWCNTs.
3.2 Protein 1. 2 mg MWCNTs with PEG functionalization are sonicated for

Conjugation (Fig. 2) 30 min in 1 mM PBS pH 7.2.
2. Activation of carboxyl groups by using equal volumes (5 mL)
of 2 mM EDC/5 mM NHS.
3. After a minute of stirring, immediately, is reacted with 1 mg/
mL of protein.
4. After 6 h reaction at 4 C, the solution is purified to remove
unconjugated protein by filtration through 0.45 μm polycar-
bonate filters and extensive washing with ultrapure water.
5. The final conjugated nanotubes are redisposed in 2 mL of
1 mM PBS pH 7.2.
3.3 Fourier The successful immobilization of the proteins is tested by FTIR.

Transform Infrared According to the spectrum obtained with the process of infrared
(FTIR) Spectroscopy spectroscopy (Fig. 2), the peak appearing at 1797 cm1 is due to
the stretching vibration of the bond C¼O of the carboxylic acids of
both nanotubes and protein. Also, the stretching vibration of the
bond C¼C at 1649 cm1 is the characteristic peak of nanotubes.
cnt cnt peg cnt peg prt

1.0 1.0 1.0
Transmittance (Normalized)
1200
1347
1391
1120
1524
1797 1524 1120
0.8 1797 1324
1524
1797
1649 0.8 0.8
1649
1649
0.6
2000 1500 1000 2000 1500 1000 2000 1500 1000
cm–1 cm–1 cm–1
Fig. 2 IR spectra of carbon nanotubes, the carbon nanotubes modified with bis(3-aminopropyl) polyethylene
glycol, and the immobilized protein with modified carbon nanotubes
The peak at 1524 cm1 is the characteristic bending vibration of N-

H bond of the associated secondary amides that are presented in
proteins and it appears only in the spectrum with the immobilized
protein. In contrast, at the spectrum of nanotubes is observed two
absorptions at 1347 cm1 and 1200 cm1 which correspond to
stretching and bending vibration of the bond C-O, respectively.
Modified nanotubes with bis(3-aminopropyl) polyethylene glycol
exhibit a peak at 1391 cm1 corresponding to a stretching vibration
bond C-O and one at 1120 cm1 which is characteristic of the C-N
bond stretching vibration of primary amines, which does not
appear in the spectrum with the protein, indicating the immobili-
zation of the modified carbon nanotubes. Finally, a peak is observed
at 1324 cm1 at the spectra of the immobilized protein due to
secondary amines.
3.4 Thermo- The successful immobilization of the proteins is also tested by

gravimetric Analysis thermogravimetric analysis (TGA). In Fig. 3 that demonstrates
(TGA) the thermogravimetric analysis of the protein is observed the reduc-
tion of the mass of nanotubes according to the increase of temper-
ature. The weight loss of nanotubes (5%) due to the removal of its
carboxylic groups and the increased weight loss (25%) of the func-
tionalized nanotubes with bis(3-aminopropyl) polyethylene indi-
cate the successful coating. Finally, the greatest weight loss (35%) is
presented by conjugated nanotubes with the protein, indicating the
successful immobilization of the majority of protein amount, con-
sidering that only 5% of the MWCNTs are oxidized.
100
80
Weight Loss (%)
60
40
CNT
20
CNT-PEG
CNT-PEG-PRT
0
0 100 200 300 400 500 600 700
Temperature (°C)
Fig. 3 Thermogravimetric analysis (TGA) of carbon nanotubes, the carbon

nanotubes modified with bis(3-aminopropyl) polyethylene glycol, and the immo-
bilized protein with modified carbon nanotubes
3.5 Zeta Potential The functionalized MWCNTs are characterized by measuring their
zeta potential. The zeta potential is evaluated by using a Malvern
Zetasizer (Nano ZS, ZEN3600, Malvern Instruments, Worcester-
shire, UK) fitted with a red laser light beam (λ ¼ 632.8 nm). The
Z-potential is calculated from the mean electrophoretic mobility
values determined by laser Doppler anemometry (LDA).
50 μL of the nanotube solutions is transferred into a particle
size cuvette without any dilution. The zeta potential measurements
are performed upon mixture with 1 mM KCl solution at volume
ratio 1:1. Each batch is analyzed in triplicate.
The surface zeta potential of immobilized MWCNTs with the
protein performs lower zeta potential (12.8 mV) (increase of the
absolute values) than the MWCNTs with bis(3-aminopropyl) poly-
ethylene (7.28 mV), which confirmed stability of construct
(Fig. 4). The lower Z-potential is due to the carboxyl groups
residing on the surface of the negatively charged immobilized
MWCNTs with the protein at physiological pH.
3.6 Cell Cultures and 1. 8 104 cells are seeded out in 24-well dishes.
In Vitro Cell Viability 2. After 8 h incubation, the medium is replaced by a fresh one
together with particular concentration of nanotubes and
immobilized proteins.
3. The cells are trypsinized, and cell viability is determined by the
trypan blue dye exclusion test.
4 Notes
1. The EDC/NHS solution activates the carboxyl groups of oxi-

dized multiwalled carbon nanotubes.
2. The surface modification of nanotubes with bis(3-aminopro-
pyl) polyethylene glycol, convert them into more hydrophilic
and non-cytotoxic. Also, PEG prevents the non-specific reac-
tions with proteins and other serum components.
Acknowledgments
We would like to acknowledge the University of Saskatchewan

College of Medicine CoMGRAD postdoctoral fellowship awarded
to Petros Kechagioglou.
A Mean (mV) Area (%) Width (mV)

Zeta Potential (mV): -7,28 Peak 1: -7,28 100,0 5,80
Zeta Deviation (mV): 5,80 Peak 2: 0,00 0,0 0,00
Conductivity (mS/cm): 0,0157 Peak 3: 0,00 0,0 0,00
Zeta Potential Distribution
600000
500000
Intensity (kcps)
400000
300000
200000
100000
0
–200 –100 0 100 200
Zeta Potential (mV)
Record 8: control
B Mean (mV) Area (%) Width (mV)

Zeta Potential (mV): -12,8 Peak 1: -12,8 100,0 5,73
Zeta Deviation (mV): 5,73 Peak 2: 0,00 0,0 0,00
Conductivity (mS/cm): 0,0190 Peak 3: 0,00 0,0 0,00
Zeta Potential Distribution
800000
700000
Intensity (kcps)
600000
500000
400000
300000
200000
100000
0
–200 –100 0 100 200
Zeta Potential (mV)
Record 5 : 2
Fig. 4 Zeta potential (mV). (A) The zeta potential of MWCNTs with bis(3-aminopropyl) polyethylene is
7.28 mV, while the immobilized MWCNTs with the protein perform lower zeta potential (12.8 mV)
References
1. Saudagar P, Dubey VK (2014) Carbon nano- multiwalled carbon nanotubes: developing a
tube based betulin formulation shows better model for cell uptake. Nano Lett
efficacy against Leishmania parasite. Parasitol 12:4370–4375
Int 63:772–776 12. Albini A, Mussi V, Parodi A, Ventura A,
2. Kang S, Herzberg M, Rodrigues DF, Elime- Principi E, Tegami S, Rocchia M,
lech M (2008) Antibacterial effects of carbon Francheschi E, Sogno I, Cammarota R (2010)
nanotubes: size does matter! Langmuir Interactions of single-wall carbon nanotubes
24:6409–6413 with endothelial cells. Nanomedicine
3. Olivi M, Zanni E, De Bellis G, Talora C, Sarto 6:277–288
MS, Palleschi C, Flahaut E, Monthioux M, 13. Jin H, Heller DA, Sharma R, Strano MS
Rapino S, Uccelletti D, Fiorito S (2013) Inhi- (2009) Size-dependent cellular uptake and
bition of microbial growth by carbon nanotube expulsion of single-walled carbon nanotubes:
networks. Nanoscale 5:9023–9029 single particle tracking and a generic uptake
4. Kang S, Mauter MS, Elimelech M (2008) model for nanoparticles. ACS Nano
Physicochemical determinants of multiwalled 3:149–158
carbon nanotube bacterial cytotoxicity. Envi- 14. Heister E, Lamprecht C, Neves V, Tı̂lmaciu C,
ron Sci Technol 42:7528–7534 Datas L, Flahaut E, Soula B, Hinterdorfer P,
5. Sumio I (1991) Helical microtubules of gra- Coley HM, Silva SRP, McFadden J (2010)
phitic carbon. Nature 354:56–58 Higher dispersion efficacy of functionalized
6. Prato M, Kostarelos K, Bianco A (2008) Func- carbon nanotubes inchemical and biological
tionalized carbon nanotubes in drug design environments. ACS Nano 4(5):2615–2626
and discovery. Acc Chem Res 41(1):60–68 15. Veronese FM, Pasut G (2005) PEGylation,
7. Heister E, Neves V, Tı̂lmaciu C, Lipert K, Sanz successful approach to drug delivery. Drug Dis-
Beltrán V, Coley H et al (2009) Triple functio- cov Today 10(21):1451
nalisation of single-walled carbon nanotubes 16. Hahn BD, Lee JM, Park DS, Choi JJ, Ryu J,
with doxorubicin, a monoclonal antibody, and Yoon WH, Lee BK, Shin DS, Kim HE (2009)
a fluorescent marker for targeted cancer ther- Mechanical and in vitro biological perfor-
apy. Carbon 47(9):2152–2160 mances of hydroxyapatite–carbon nanotube
8. Ladeira MS, Andrade VA, Gomes ERM et al composite coatings deposited on Ti by aerosol
(2010) Highly efficient siRNA delivery system deposition. Acta Biomater 5(8):3205–3214
into human and murine cells using single-wall 17. Khan AS, Hussain AN, Sidra L, Sarfraz Z,
carbon nanotubes. Nanotechnology Khalid H, Khan M, Manzoor F, Shahzadi L,
21:385101 Yar M, Rehman IU (2017) Fabrication and
9. Pantarotto D, Singh R, McCarthy D, in vivo evaluation of hydroxyapatite/carbon
Erhardt M, Briand JP, Prato M, Kostarelos K, nanotube electrospun fibers for biomedical/
Bianco A (2004) Functionalized carbon nano- dental application. Mater Sci Eng C
tubes for plasmid DNA gene delivery. Angew 80:387–396
Chem Int Ed Engl 43(39):5242–5246 18. Marrs B, Andrews R, Rantell T, Pienkowski D
10. Kostarelos K, Lacerda L, Pastorin G, Wu W, (2006) Augmentation of acrylic bone cement
Wieckowski S, Luangsivilay J et al (2007) Cel- with multiwall carbon nanotubes. J Biomed
lular uptake of functionalized carbon nano- Mater Res A 77(2):269–276
tubes is independent of functional group and 19. Tsukasa A, Keiko N, Motohiro U, Fumio W
cell type. Nat Nanotechnol 2:108–113 (2009) Modification of the dentin surface by
11. Mu QX, Broughton DL, Yan B (2009) Endo- using carbon nanotubes. Biomed Mater Eng
somal leakage and nuclear translocation of 19(2–3):179–185
Chapter 13
Peptide-Mediated Biomimetic Regrowth of Human

Enamel In Situ
Kaushik Mukherjee, Qichao Ruan, and Janet Moradian-Oldak
Abstract
Mimicking the dynamics of mineral loss and gain involved in dental caries formation can help us evaluate
and compare the mineralization efficacy of different treatment agents used in enamel remineralization.
Here, we offer an abridged study design outlining the preparation of tooth samples, creation of artificial
dental lesions, application of a peptide, and characterization of the regrown enamel-like mineral layer.
Key words Tooth enamel, Demineralization, Remineralization, Regrowth, Artificial saliva, Apatite,
Leucine-rich amelogenin peptide (LRAP)
1 Introduction
Dental caries is one of the most ubiquitous childhood disease pro-

cesses whereby net mineral loss from the dental hard tissues com-
promises tooth structure and function [1]. Mimicking dental lesions
through in vitro and in situ studies can help to improve the under-
standing of the dynamics of mineral loss and gain. A variety of dental
lesions and pH-cycling model systems have been established to
mimic and investigate the mechanistic aspects of caries formation
[2]. Such studies need to be carefully designed in a clean,
contamination-free environment and run for a particular duration
of time to achieve the best outcomes. Current remineralizing stra-
tegies aim to revolutionize the treatment practices in operative
dentistry by rebuilding superficial enamel [3, 4]. This bioinspired
process is achieved through a critical understanding of the func-
tional role of natural enamel-forming proteins (such as amelogenin)
in mediating apatite mineralization in vitro [5–7]. The ultimate
objective of these studies is to develop a next-generation dental
material that can impede the progression of tooth decay, restore
and replace lost enamel tissue, and prolong the integrity of the
tooth. In this chapter we provide a simplified protocol to conduct
129
130 Kaushik Mukherjee et al.
a demineralization-remineralization experiment in evaluating the

peptide-mediated remineralization of artificial dental lesions and
characterizing the orientation, formation, and composition of the
newly formed enamel-like apatite layer. We chose to present the case
of leucine-rich amelogenin peptide (LRAP) as a model peptide for
the enamel remineralization experiments [4, 8, 9].
2 Materials
2.1 For Tooth 1. Sound human 3rd molars.

Selection, Storage, 2. Ethanol, tweezers, dental probes, and scaler for cleaning
and Preparation molars and removing excess soft tissue.
3. Phosphate-buffered saline (PBS, pH 7.4) as tooth storage
media plus.
4. 0.002% sodium azide (microbial inhibitor).
5. Ultrasonic bath or cleaner.
6. 400–4000 grit silicon carbide papers.
7. Nylon adhesive back discs.
8. 0.25 μm diamond or colloidal silica suspension.
9. Water-resistant nail varnish.
2.2 For 1. Demineralization buffer: 2 mM CaCl2·2H2O, 2 mM KH2PO4,

Demineralization/ 50 mM sodium acetate, and 0.879 mL acetic acid adjusted at
Remineralization pH 4.6.
Buffer and Peptide 2. pH probe.
Sample Preparation 3. Artificial saliva: 1.2 mM CaCl2·2H2O, 0.72 mM K2HPO4,
16 mM KCl, 0.2 mM MgCl2·6H2O, 50 mM HEPES, and
4.5 mM NH4Cl in DDW adjusted to pH 7.2.
4. Fluoride (1 ppm) in the form of NaF.
5. Millex-GV, 0.22 μm disposable filter unit.
6. Peptide (0.2 mg/mL) and 2% (w/v) chitosan adjusted to
pH 6.5.
2.3 For Embedding 1. Self-cure denture base clear powder (G10–0300)/liquid

Tooth Samples in (G10–0136) (Pearson Dental Co.).
Resin 2. Cement spatula.
3 Methods
3.1 Tooth Selection 1. Collect sound human molars with no detectable signs of
and Storage (1) caries, (2) fillings, (3) discoloration (due to fluorosis, ame-
logenesis imperfecta, etc.), (4) cracks/fracture, or (5) marked
Enamel Biomimetics 131
erosive or abrasive wear according to a protocol approved by

the responsible Institutional Review Board.
2. Remove excess soft tissue deposits and calculus by scrupulous
cleaning, scaling, and pulp extirpation (using dental probe).
Rinse teeth in 70% ethanol, and place in distilled water for a
20-min sonication. Wash the cleaned tooth samples thor-
oughly, and store them in tooth storage media composed of
diluted phosphate-buffered saline (PBS, pH 7.4) with 0.002%
sodium azide used as a microbial inhibitor at 4 C [10]. Alter-
natively, Hank’s Balanced Salt Solution (HBSS) can also be
prepared as a storage medium [11] (see Notes 1 and 2).
3.2 Demineralized 1. Remove the tooth roots (decoronate), and section the remain-
Enamel Lesion ing crown buccolingually and longitudinally into 2-mm-thick
Preparation slices using a water-cooled, slow-speed diamond saw.
2. Polish the tooth slices with a sequential series of wet 400–4000
grit silicon carbide papers and nylon adhesive back discs with
0.25 μm diamond or colloidal silica suspension. Rinse the
polished slices thoroughly with distilled water (DDW) three
times, sonicate in a water bath for 5 min, rinse again, and allow
to air-dry.
3. Paint the surfaces of the tooth samples with two layers of clear
acid-resistant nail varnish, leaving only an exposed window of
dimensions 3 2mm2. Let the samples air-dry at room tem-
perature (RT) for 3–4 h (see Note 3).
4. Place the varnish-coated enamel specimens in a demineraliza-
tion solution (2 mM CaCl2, 2 mM KH2PO4, 50 mM sodium
acetate, 0.879 mL acetic acid) adjusted to pH 4.6 using 1 M
HCl at 37 C for 2 h (5 teeth sections/beaker in 50 mL
solution).
3.2.1 Preparation of 1. Peptides used for enamel remineralization can be easily synthe-
Peptide sized commercially using solid phase peptide synthesis [12].
Posttranslational modifications such as phosphorylation can be
incorporated in the amino acid sequence. In the case of leucine-
rich amelogenin peptide (LRAP, 59 amino acids length), a
phosphoryl group is added at Ser16. Use high-performance
liquid chromatography (HPLC) and mass spectrometry for
peptide purification and determination. The peptides to be
used should be ~> 95% pure (see Note 4).
2. Weigh the peptide sample and dissolve in ultrapure distilled
water to yield a stock solution of 2mg/mL, and centrifuge at
10,000 rpm (8944 g) for 5 min. Place the stock solution in a
slow shaker for 4 h at 4 C, and then divide into aliquots of
100 μL/tube.
3. Lyophilize the aliquots for 12 h to yield a final peptide concen-
tration of 200 μg per tube (optimized peptide concentration).
4. Prior to the remineralization cycle, dissolve LRAP (200 μg) in

filtered deionized water (DDW) (960 μL) at room temperature
with Na2HPO4 (15 μL, 0.1 M) and CaCl2 (25 μL, 0.1 M) to
yield a solution of concentration with 0.2 mg/mL LRAP.
Adjust the final pH value to ~6.5–7 (close to salivary pH)
with 1 M KOH (dilute to 100 mM NaOH if no buffer in
peptide solution to avoid sudden changes in the pH value).
Centrifuge the sample (10,000 rpm, 5 min) just prior to use.
3.2.2 Preparation of 1. Peptides can also be incorporated in a chitosan-based hydrogel

LRAP-Chitosan (CS-LRAP) and applied on the dental lesions [3, 4]. To prepare chitosan
Hydrogel stock solution (CS), dissolve 2% (w/v) chitosan (medium
molecular weight, 75–85% deacetylated) in a 1% (v/v) acetic
acid solution followed by stirring at 80 C overnight.
2. After cooling the solution to room temperature, pass it
through a 0.45 μm filter. Adjust the pH value to 6.5 by adding
1 M NaOH solution.
3. Add LRAP to the chitosan gel: Mix chitosan (medium molec-
ular weight, 75–85% deacetylated, Sigma-Aldrich) solution
(960 mL, 2% m/v), Na2HPO4 (15 μL, 0.1 M), CaCl2
(25 μL, 0.1 M), and LRAP (200 μg), followed by stirring at
room temperature overnight. Adjust the pH value to 6.5 (~pI
of chitosan) by adding 1 M NaOH solution.
3.3 Remineralization 1. Prepare artificial saliva solution (1 L) with a final concentration

Media and Biomimetic of 1.2 mM CaCl2·2H2O, 0.72 mM K2HPO4, 16 mM KCl,
Enamel Regrowth 0.2 mM MgCl2·6H2O, 50 mM HEPES, and 4.5 mM NH4Cl
in DDW. Adjust to pH 7.2 using 1 M KOH. Stir the solution
for 10 min to ensure all the ingredients are dissolved. Filter the
stock solution (Millex-GV, 0.22 μm filter unit) three times
prior to use.
2. To remineralize in artificial saliva: Apply 20–30 μL of prepared
protein solution or CS-LRAP hydrogel (see Subheadings 3.2.2)
to each enamel window, and let it dry in the desiccator for
10 min at room temperature.
l Immerse the protein-treated tooth slice in 5 mL of artificial
saliva each at 37 C for a predetermined number of days.
Replenish the artificial saliva every 24 h along with 1 ppm F
(see Note 5) (Fig. 1).
l After incubation (end of remineralization cycle), sonicate
the tooth slices in a water bath for 10 min to remove any
surface debris or loosely bound crystals, gently rinse with
deionized water, and air-dry for further assessment using
techniques such as X-ray diffraction, nanoindentation, and
scanning electron microscopy. Store the remaining samples
in DDW at 4 C for future use (see Note 6).
Fig. 1 An image demonstrating LRAP-CS hydrogel-coated tooth slice after being

immersed in artificial saliva. Note the gel remains adhered (arrow) to the tooth
surface forming a shell-like covering
3.4 Assessment and XRD is a powerful method employed for crystallographic analysis.
Characterization of
1. For our studies, we use a diffractometer with monochroma-
Biomimetic Enamel-
tized Cu(Kα) radiation (λ ¼ 0.154 nm) at 40 kV and 44 mA with
Like Layer a sampling step size of 0.08 and 2θ range of 5–65 to analyze
3.4.1 X-Ray Diffraction the crystal orientation and mineral phase of the newly formed
(XRD) crystals.
2. To extrapolate additional details from the XRD data:
(a) Index the diffraction peaks referring to the standard
JCDPS file (#09–0432) using MDI JADE 6.
(b) Use the R-value (ratio of intensities of 002 and 211) to
find the orientation degree of the HAp crystals (higher
value indicates preferred orientation).
(c) For enamel powder samples, use the crystalline index
(CI)XRD, degree of crystallinity, and the Debye-Scherrer
equation (correlates the size of crystallites with peak
broadening) to determine the average crystallite size, per-
fection, and ordering in a sample [13–15].
The degree of crystallinity can be calculated using diffrac-
tion peaks (112) or (211) and (300) of HAp with the following
equation:

X c ¼ I 300 V 112=300 =I 300 100 ½%
where Xc is defined as the fraction of crystalline phase, I300 is
the (300) diffraction peak intensity, and V112/300 is the inten-
sity of the trough between (112) and (300) diffraction peaks of
HAp. The range of 2θ where the peaks of interest fall is graphed
separately and then fitted using a Gaussian/Lorentzian
function.
The Debye-Scherrer equation:

T ðhklÞ ¼ 0:9λ cos =B cos θðhklÞ
where λ is the wavelength of the monochromatic X-ray beam,
B is the full width at half maximum (FWHM) of the peak at the
maximum intensity, θ(hkl) is the peak diffraction angle that
satisfies Bragg’s law for the (hkl) plane, and t(hkl) is the crystal-
lite size. The (002) reflection peak at 25.9 from the XRD
pattern of HAp can be used to calculate the nano-HAp
crystallite size.
3.4.2 Scanning Electron To explore the morphology of sound, demineralized, and reminer-
Microscopy alized enamel surfaces and to investigate the interface between
synthetic and native enamel (on the macro-microscopic scale),
scanning electron microscopy equipped with an energy-dispersive
detector (EDAX) can be used at different levels of magnification:
1. Mount the tooth specimens on aluminum stubs with a carbon
tape, sputter coat with Au/Pt for ~30 s (~5–10 nm thick
coating), and observe under an accelerating voltage of 10 kV.
Both top-down and side views of the sectioned tooth samples
can be observed under SEM after the remineralization cycle
(Fig. 2).
2. To observe the cross section of the newly formed layers, the
tooth slices can be embedded in resin:
(a) Fill a plastic mold with a thin layer of self-curing resin
polymer, and moisten with a drop of the monomer (see
Note 7).
Fig. 2 SEM images of the newly grown layer after remineralization in LRAP-CS hydrogel for 3 days (a) top view
and (b) cross-sectional view. (arrow shows the newly grown layer)
Fig. 3 Typical EDXS curve for healthy enamel (a) and the CS-LRAP-treated enamel surface after 7 days of
remineralization (b). The chemical composition of the regenerated apatite layer is similar to that of native
enamel
(b) Place each tooth section parallel to the mold space to

guarantee the precision of the section, and pour the resin
into the remaining space using the salt and pepper method.
Allow the resin to cure and harden for up to 2 h in RT.
(c) Extract the resin block from the mold, and make a longi-
tudinal cut through the window using a water-cooled
diamond saw advancing at low speed.
(d) Polish the cross sections with wet grid papers and nylon
cloth (as described above) using gentle force, rinse in
ethanol, sonicate for 2 min in distilled water, rinse thor-
oughly three times, and blast with air to dry the sample,
and remove any remaining polishing suspension particles.
Repeat step 1 above to image using SEM.
3. Simultaneously, energy-dispersive X-ray spectroscopy
(EDX/EDS/EDXS) can be used for the semiquantitative ele-
mental analysis or chemical characterization of healthy, demi-
neralized, and remineralized tooth samples (Ca/P molar ratio
and weight % of elements such as Ca, P, F, Na, Mg, O, C, etc.
can be calculated) (Fig. 3a, b).
3.4.3 Microhardness 1. The microhardness of repaired enamel is an important indica-

Tests tor of the durability and strength of the regenerated apatite
layer. After the remineralization cycle, rinse and sonicate the
treated tooth slices for 10 min to ensure all the loosely bound
crystals are effectively removed from the enamel surface. The
two most routinely used surface microhardness (SMH) meth-
ods in tooth enamel studies are (1) Knoop hardness (HK) and
(2) Vickers hardness (HV) tests [16]. For HK, only the longer
diagonal is measured, and the hardness is calculated by dividing

the projected area of indent with the applied force (kgf/mm2).
For HV, both the diagonals are measured, and the average is
calculated to determine the Vickers pyramid number.
2. Measure surface microhardness (SMH) with a hardness tester
using a load force in the range 25–100 g force and 10–20s
dwell time, both before and after the remineralization cycle (see
Note 8). For each sample, make 6–10 indentations on the
surface spaced 100 μm apart (should be ~2.5 times the indent
diagonal). Following the treatment, calculate the degree of
hardness recovery as %SMHR ¼ 100 (SMH2–SMH1)/
(SMH0–SMH1), comparing with the microhardness of healthy
and demineralized enamel. Here, SMH0 is the surface hardness
at baseline (healthy enamel); SMH1 is the surface hardness of
the acid-treated demineralized lesion, and SMH2 is the surface
hardness after treatment with peptide in artificial saliva. Use the
same calibrated machine for before and after treatment mea-
surements (see Note 9). Calculate the average microhardness
value for at least five specimens per sample group, and compare
the differences in the HK or HV using two-way ANOVA
followed by a Tukey test. For thin remineralized coatings,
nanoindentation is a more accurate method of assessing
changes in the mechanical properties.
3. Nanoindentation measurement: The nanoindentation tech-
nique may be performed to study the mechanical behavior
and reliability of dental enamel more accurately. Mount and
stabilize the samples on acrylic slabs [17]. Use a Berkovich
diamond indentation tip (with a curvature less than 100 nm)
to make indents on the sample surface (25 indents/tooth
section). A continuous stiffness measurement (CSM) is used
to measure the hardness (strength) and the elastic modulus
(stiffness) of the regrown apatite layer. Set the following para-
meters under CSM mode: a target constant strain-rate (CSR)
of 0.05s-1, measuring depth range 500nm–1000nm (1000nm
depth used in our experiments), and keep the distance between
the indents to 100μm to prevent interferences (see Note 10).
4 Notes
1. Take care to maintain a clean, sterile working environment,

exercise lab safety measures, and wear personal protective
equipment (mask, lab coat, safety glasses).
2. Replenish the tooth storage media every 1–2 months as pro-
longed storage (~12 months) may significantly decrease the
microhardness of dental tissues [18]. The significant effect of
fluoride (1 ppm) in promoting enamel remineralization is
important [19]. Hence, studies adding fluoride to the peptide

solution for in vitro enamel remineralization should have a
control to demonstrate the effect of fluoride only and
peptide-fluoride combination.
3. The enamel window should be made on the same location for
all the tooth specimens (e.g., buccal cusp tips of molars away
from the DEJ).
4. LRAP can be replaced by full-length amelogenin or other
peptides for enamel mineralization studies in vitro [3, 20–21].
5. The number of treatment applications and the duration of the
remineralization can be tailored to suit the individual aim and
requirement of the experiment. Ensure the inclusion of a vali-
dated positive control in all such experiments.
6. A range of caries remineralization models can be developed to
investigate different aspects of the caries process (such as
non-cavitated lesions, root caries, role of biofilm, etc.) [22].
7. Because the resin is a skin irritant, do not handle it with bare
hands, avoid skin contact, and wear protective eyeglasses.
8. The indentation load for the microhardness test can be per-
formed using 1–1000 g and with various loading times. The
preferred range has been mentioned in the text. However, the
microhardness results on enamel may not be constant at very
low loads [16].
9. In addition to microhardness tests, scratch test may be per-
formed under standard conditions (loading rate 50 N/min and
scratching speed 3 mm/min) to demonstrate the adhesion
between synthetic and native enamel structure [23].
10. The penetration depth of the indenter tip may be adjusted
depending on the thickness of the regrown layer. For thin
remineralized coatings, the depth should be 600 nm to mini-
mize the substrate effects arising from the underlying sound
enamel. Increasing the penetration depth range on healthy
enamel surface from 100 to 2000 nm may drastically decrease
the hardness and modulus by almost 30% [24].
Acknowledgments
This research was supported by NIH-NIDCR R01 grants

DE-13414 and DE-020099 and the USC Coulter Translational
Partnership Program.
References
1. Selwitz RH, Ismail AI, Pitts NB (2007) Dental 13. Person A, Bocherens H, Saliège JF, Paris F,
caries. Lancet 369(9555):51–59 Zeitoun V, Gérard M (1995) Early diagenetic
2. Buzalaf MA, Hannas AR, Magalhães AC, evolution of bone phosphate: an X-ray diffrac-
Rios D, Honório HM, Delbem AC (2010) tometry analysis. J Archaeol Sci 22(2):211–221
pH-cycling models for in vitro evaluation of 14. Poralan Jr. GM, Gambe JE, Alcantara EM,
the efficacy of fluoridated dentifrices for caries Vequizo RM. X-ray diffraction and infrared
control: strengths and limitations. J Appl Oral spectroscopy analyses on the crystallinity of
Sci 18(4):316–334 engineered biological hydroxyapatite for medi-
3. Ruan Q, Zhang Y, Yang X, Nutt S, Moradian- cal application. In IOP conference series: mate-
Oldak J (2013) An amelogenin–chitosan rials science and engineering 79, 1, 012028).
matrix promotes assembly of an enamel-like IOP Publishing Bristol. 2015
layer with a dense interface. Acta Biomater 9 15. Klug HP, Alexander LE (1974) X-ray diffrac-
(7):7289–7297 tion procedures for polycrystalline and amor-
4. Mukherjee K, Ruan Q, Liberman D, White SN, phous materials, 2nd edn. Wiley, New
Moradian-Oldak J (2016) 2016. Repairing York-London, p 689
human tooth enamel with leucine-rich amelo- 16. Chuenarrom C, Benjakul P, Daosodsai P
genin peptide–chitosan hydrogel. J Mater Res (2009) Effect of indentation load and time on
31(5):556–563 knoop and vickers microhardness tests for
5. Ruan Q, Moradian-Oldak J (2015) Amelo- enamel and dentin. Mater Res 12(4):473–476
genin and enamel biomimetics. J Mater Chem 17. Chung HY, Huang KC (2013) Effects of pep-
3(16):3112–3129 tide concentration on remineralization of
6. Beniash E, Simmer JP, Margolis HC (2005) eroded enamel. J Mech Behav Biomed Mater
The effect of recombinant mouse amelogenins 28:213–221
on the formation and organization of hydroxy- 18. Aydın B, Pamir T, Baltaci A, Orman MN, Turk
apatite crystals in vitro. J Struct Biol 149 T (2015) Effect of storage solutions on micro-
(2):182–190 hardness of crown enamel and dentin. Eur J
7. Fang PA, Conway JF, Margolis HC, Simmer Dent 9(2):262
JP, Beniash E (2011) Hierarchical self-assembly 19. Ten Cate JM, Featherstone JDB (1991) Mech-
of amelogenin and the regulation of biominer- anistic aspects of the interactions between fluo-
alization at the nanoscale. Proc Natl Acad Sci ride and dental enamel. Crit Rev Oral Biol Med
108(34):14097–14102 2(3):283–296
8. Le Norcy E, Kwak SY, Wiedemann-Bidlack FB, 20. Kirkham J, Firth A, Vernals D, Boden N,
Beniash E, Yamakoshi Y, Simmer JP, Margolis Robinson C, Shore RC, Brookes SJ, Aggeli A
HC (2011) Leucine-rich amelogenin peptides (2007) Self-assembling peptide scaffolds pro-
regulate mineralization in vitro. J Dent Res 90 mote enamel remineralization. J Dent Res 86
(9):1091–1097 (5):426–430
9. Shafiei F, Hossein BG, Farajollahi MM, 21. Yang Y, Lv XP, Shi W, Li JY, Li DX, Zhou XD,
Fathollah M, Marjan B, Tahereh JK (2015) Zhang LL (2014) 8DSS-promoted reminerali-
Leucine-rich amelogenin peptide (LRAP) as a zation of initial enamel caries in vitro. J Dent
surface primer for biomimetic remineralization Res 93(5):520–524
of superficial enamel defects: An in vitro study. 22. Cochrane NJ, Zero DT, Reynolds EC (2012)
Scanning 37(3):179–185 Remineralization models. Adv Dent Res 24
10. Reed R, Xu C, Liu Y, Gorski JP, Wang Y, (2):129–132
Walker MP (2015) Radiotherapy effect on 23. Wu D, Yang J, Li J, Chen L, Tang B, Chen X,
nano-mechanical properties and chemical com- Wu W, Li J (2013) Hydroxyapatite-anchored
position of enamel and dentine. Arch Oral Biol dendrimer for in situ remineralization of
60(5):690–697 human tooth enamel. Biomaterials 34
11. Habelitz S, Marshall GW, Balooch M, Marshall (21):5036–5047
SJ (2002) Nanoindentation and storage of 24. Zhou J, Hsiung LL (2007) Depth-dependent
teeth. J Biomech 35(7):995–998 mechanical properties of enamel by nanoinden-
12. Amblard M, Fehrentz JA, Martinez J, Subra G tation. J Biomed Mater Res Part A 81
(2006) Methods and protocols of modern solid (1):66–74
phase peptide synthesis. Mol Biotechnol 33
(3):239–254
Chapter 14
Bioengineering Tooth Bud Constructs Using GelMA Hydrogel

Elizabeth E. Smith and Pamela C. Yelick
Abstract
Bioengineered dental tissues and whole teeth that exhibit features and properties of natural teeth can
functionally surpass currently used artificial dental implants. However, no biologically based alternatives
currently exist for clinical applications in dentistry. Here, we describe a newly established bioengineered
tooth bud model for eventual applications in clinical dentistry. We also describe methods to fabricate and
analyze bioengineered tooth tissues, including cell isolation, in vivo implantation, and post-harvest
analyses.
Key words Tooth tissue engineering, Hydrogel scaffolds, Primary dental cell culture, Odontogenesis
1 Introduction
Currently, artificial dental implants are commonly used to replace

human teeth that have been lost due to trauma, cancer resection, or
birth defects. Although this therapy has been proven to be a suc-
cessful approach to replace lost teeth in many individuals, a variety
of common patient complications can occur, such as gingival tissue
and bone loss, peri-implantitis, and even implant failure [1–4]. It
has been suggested that biological-based living tooth replacements
can overcome many of the complications observed in synthetic
dental implant therapy [5–7]. One area of current dental research
aims to fabricate bioengineered whole teeth by using natural tooth
development as a guide [6, 8, 9]. Using this approach, tissue
engineering strategies were used to identify suitable cell popula-
tions, and scaffold materials and designs, to support dental cell
attachment, viability, proliferation, and differentiation [6, 8, 10].
It was shown that gelatin methacrylate (GelMA) hydrogel scaffolds
can be tuned to mimic certain features of immature natural dental
tissues and to support the differentiation of encapsulated dental
139
140 Elizabeth E. Smith and Pamela C. Yelick
cells, including elaboration of mineralized dental tissue formation

[9]. Here, we describe in detail the fabrication and analytical meth-
ods used to create and validate a novel bioengineered 3D GelMA
tooth bud model.
2 Materials
Prepare all solutions using ultrapure water (prepared by purifying

deionized water, to attain a sensitivity of 18 MΩ-cm at 25 C) and
analytical grade reagents. Prepare and store all reagents at room
temperature (unless indicated otherwise). Diligently follow all
waste disposal regulations when disposing waste materials. We do
not add sodium azide to reagents. All tissue culture solutions are
sterilized by autoclaving and/or filtration through 22 micron filters.
2.1 Primary Dental 1. Collagenase/dispase solution: Prepare 5–10 min prior to use.
Cell Isolation, In Vitro Dissolve 20 mg collagenase type II and 10 mg dispase in
Expansion, and 50 mL PBS.
Cryopreservation 2. Dental mesenchymal (DM) culture media (for cells/tissues
harvested from human or porcine pulp organ): Advanced
DMEM/F12 media supplemented with 10% FBS, 25 μg/mL
ascorbic acid, 1 PSA, and 1 Glutamax.
3. Dental epithelial (DE) culture media (for cells/tissues har-
vested from human or porcine enamel organ): LHC-8 media
supplemented with 10% FBS, 0.5 μg/mL epinephrine, and
1 PSA.
2.2 Preparation and 1. Bioengineered Tooth Bud (BTB) Culture Media: Combine
Fabrication of GelMA- 250 mL advanced DMEM/F12 media with 250 mL LHC-8
Encapsulated Dental media, supplemented with 10% FBS, 0.5 μg/mL epinephrine,
Cell Constructs 100 nM dexamethasone, 10 mM beta-glycerophosphate,
50 μg/mL ascorbic acid, and 1 PSA.
2. Lyophilized Gelatin Methacrylate (GelMA): The GelMA lyo-
philizate used to establish these methods was a generous gift
from Dr. Ali Khademhosseini. The synthesis of GelMA lyophi-
lizate has been thoroughly described [11, 12].
2.3 Live/Dead 1. Live/Dead Staining Kit: Live/Dead Viability/Cytotoxicity Kit

Analysis of In Vitro (Molecular Probes). Thaw reagents to 37 C prior to use.
Cultured GelMA Tooth
Bud Constructs
2.4 Subcutaneous 1. Rat host: Immunocompromised 5-month-old female Rowett

Implantation, Harvest, Nude rats supplied by Charles River Laboratories.
and Fixation of GelMA
Tooth Bud Constructs
Novel Bioengineered Tooth Bud Model 141
2.5 Processing and 1. Decalcification Solution (22.5% formic acid + 10% sodium
Analyses of GelMA citrate): Prepare 45% formic acid in DI H2O by slowly adding
Tooth Bud Constructs 225 mL of 98% formic acid to 275 mL of DI H2O. Next
prepare 20% sodium citrate by slowly adding 100 g of sodium
2.5.1 Decalcification and citrate to 400 mL of DI H2O, and then add H2O to bring to
Processing of GelMA Tooth 500 mL volume. Carefully combine 500 mL of 20% sodium
Bud Constructs citrate to 500 mL of 45% formic acid.
2. Saturated Ammonium Oxalate: Fully dissolve 10 g of ammo-
nium oxalate in 100 mL of H2O. Slowly add additional ammo-
nium oxalate until saturated, and then filter.
2.5.2 Paraffin Embedding 1. Molten paraffin: Heat paraffin to 65 C manually or via auto-
and Sectioning matic Thermo Shandon Citadel 2000 Tissue Processor
(Thermo Fisher Scientific) and/or Microm EC 350-1 Paraffin
Embedding Center (Microm International GmbH).
2. 65 C oven or hot plate to hold molten paraffin and samples.
3. Sectioning microtome: Microm HM 355S (Thermo Fisher
Scientific).
2.5.3 Hematoxylin and 1. Mayor’s hematoxylin working solution: Dissolve 5 g of hema-

Eosin Staining toxylin in 800 mL H2O. Bring to a boil; then cool to room
temperature and cover with aluminum foil to keep in dark. Stir
overnight with a magnetic stirrer over a stir plate at room
temperature. Add 1 g of sodium iodate and 50 g aluminum
ammonium sulfate. Once dissolved, use a pipette to add
200 mL glycerol. Next, add 8 mL glacial acetic acid and mix
for 2 h. Filter and adjust pH to 2.4–2.8.
2. Diluted hydrochloric acid: Carefully add 2 mL of 37% hydro-
chloric acid to 250 mL H2O.
3. Ammonium hydroxide working solution: Add 4 mL (28–30%)
ammonium hydroxide to 250 mL H2O.
4. Eosin stock solution: Dissolve 1 g of eosin Y in DI H2O to total
volume 100 mL.
5. Phloxine B stock: Dissolve 1 g phloxine B in 100 mL H2O total
volume.
6. Eosin working solution: Carefully combine 100 mL eosin stock
solution, 10 mL phloxine B, 780 mL 95% ethanol, and 8 mL
glacial acetic acid. Mix well and adjust pH to 4.5.
2.5.4 Immunohisto- 1. 3% H2O2: Add 25 mL of 30% H2O2 to 225 mL 100% metha-

chemical Staining Using nol, keeping solutions in the dark.
Vectastain Kit and DAB 2. TE buffer (10 mM Tris +1 mM EDTA): Dissolve 1.210 g
Detection Trizma and 0.372 g EDTA into 900 mL H2O. Bring to
1000 mL total volume with H2O.
3. Blocking solution (10% donkey serum): Add 0.5 mL of 100%

normal donkey serum to 4.5 mL PBS, final volume 5.0 mL.
4. 2% donkey serum: Mix 2 mL of 10% donkey serum with 8 mL
of PBS, total volume 10 mL.
5. ABC reagent (Vectastain ABC Kit): In the dark, add 2.5 mL of
PBS to a 15 mL conical tube covered in aluminum foil. Add
1 drop of Reagent A and mix well. Then add 1 drop of Reagent
B and mix well. Allow mixing for 45 min prior to use.
6. DAB substrate (Sigma Fast DAB Tablets): Prepare within
5–10 min of use. In the dark, add 5 mL of distilled water in a
15 mL conical tube covered in aluminum foil. Add one urea
tablet and mix until dissolved. Add one DAB tablet and
mix well.
7. 0.2% Fast Green: Dissolve 0.2 g Fast Green into 500 mL H2O
at 37 C for 1 h and then filter.
2.5.5 Immunofluore- 1. 10 mM Citric Acid Solution: Dissolve 1.07 g of citric acid

scence Staining monobasic in 500 mL H2O and adjust to pH 6.0. Store at
4 C up to 1 week.
2. Blocking Solution (5% Bovine Serum Albumin, BSA): Dissolve
0.5 g in 10 mL H2O.
3. 1% BSA: Mix 2 mL of 5% BSA with 8 mL of H2O.
3 Methods
3.1 Primary Dental 1. Isolate unerupted molar tooth buds from harvested mandibu-
Cell Isolation, In Vitro lar jaws of 3–5-month-old pigs, and place in PBS. Briefly
Expansion, and identify depression in the jaw indicating location of unerupted
Cryopreservation of molar tooth bud. Use a hammer and chisel to create a window
Primary Dental Cells in the mandibular bone, and remove bone flap. Remove uner-
upted tooth bud using sterile forceps. Wash isolated tooth bud
in 1 PBS 3, 5 min each. Under aseptic conditions, place
washed tooth bud in a petri dish containing PBS. Use a dis-
secting microscope to dissect out the enamel organ, any miner-
alized tooth cusps if present, and the dental pulp organ, by
cutting at the cervical margin using a #10 scalpel and forceps.
Place the dissected enamel organ in a separate petri dish con-
taining PBS. Do not allow tissues to become dry. Place the
harvested pulp organ in a clean petri dish containing PBS;
separate from the enamel organ. At all times, keep the isolated
enamel organ and pulp organ tissues and cells separate and
hydrated in PBS.
2. Separately mince enamel organ, and pulp organ tissues using
2 #10 blades to obtain approximately 1–2 mm3-sized pieces.
Place minced tissues in a 50 mL conical tube with PBS, and
invert tubes several times to wash. Allow the minced tissues to

settle to the bottom of the tube, and then gently remove PBS.
Repeat wash 2.
3. After last wash, remove PBS and add 25 mL of collagenase/
dispase solution per tissue group. Incubate minced tissue tubes
for 30 min rotating or rocking slowly at 37 C.
4. Gently titurate the tissue suspension first with a 25 mL pipette
for 5 min and then with a 10 mL pipette for an additional
5 min.
5. Filter the cell/tissue suspension through a 40 μm cell strainer
by gravity, into a new sterile 50 mL conical tube.
6. Collect tissue from the strainer by rinsing, and place in a T175
flask with 25 mL of corresponding cell/tissue growth media
(DE media for enamel organ and DM media for pulp organ
tissues) (see Subheading 2).
7. Centrifuge the filtered cell suspension for 5 min at 1144 g at
room temperature. Remove supernatant and add 30 mL of the
appropriate growth media. Centrifuge again for 5 min at
1500 rpm at RT. Remove supernatant, and wash again with
30 mL of growth media. Centrifuge for 5 min at 1500 rpm at
RT. Finally, resuspend cells in 40 mL of appropriate growth
media.
8. Use trypan blue stain to count cells. Add 25 μL of trypan blue
to 25 μL of cell suspension. Use a hemocytometer or a cell
counter (e.g., Countess Automated Cell Counter, Invitrogen)
to count cells.
9. To culture the pulp organ cells, add two to three million cells to
each T75 culture flask or five million cells to each T175 culture
flask. Next add 15 mL of dental mesenchymal growth media to
T75 flasks and 25 mL to T175 flasks.
10. To culture the enamel organ cells, add 10 million cells per T75
flask or 20 million cells per T175 flask. Add 15 mL of dental
epithelial growth media to T75 flasks and 25 mL to T175
flasks.
11. Culture tissues and cells separately in a humidified incubator
with 5% CO2 at 37 C. Monitor cell growth daily. Change
medium when at least 10% of cell attachment can be seen.
Then change medium every 2–3 days.
12. Passage or cryopreserve the cells when the cells reach ~80–95%
confluency.
3.2 Preparation and 1. Prepare 20% photoinitiator (PI) (Irgacure2959). Fully dissolve
Fabrication GelMA- 0.2 g PI in 100% methanol in an amber microcentrifuge tube.
Encapsulated Dental Adjust volume to 1 mL with methanol. Keep in the dark and at
Cell Constructs room temperature prior to use.
Fig. 1 Fabrication and analysis of in vivo bioengineered GelMA tooth bud constructs. (a) Cultured dental cells
are harvested, pelleted, and resuspended in GelMA hydrogel (a1). The GelMA/cell solution is then pipetted into
PDMS ring molds that are placed within a 24-well plate (a2). UV exposure is used to photocrosslink the GelMA/
cell solution (a3). The PDMS molds are then removed and media is added for in vitro culture (a4). (b) Four
incisions are made on the backs of immunocompromised rats (b1) to make subcutaneous pockets. The GelMA
tooth bud constructs are carefully placed within the pockets (b2). To harvest, the skin and subcutaneous tissue
are dissected to reveal the tooth bud constructs (b3). Formalin-fixed paraffin-embedded sections can be
stained with various histological stains such as H&E (c). IF and IHC can be used to investigate the expression of
various proteins including amelogenin (AM, d) and dentin sialophosphoprotein (DSPP, e) which will be
detected by brown staining in contrast to a negative control (f). Scale bar: c–d 200 μm, insets 50 μm
2. Prepare 3–5% GelMA. Measure out appropriate amount of

lyophilized GelMA, and place in a 50 mL conical tube. In a
biosafety hood, dissolve GelMA with warmed DMEM/F12
media (see Note 1). Maintain in the dark at 60 C until fully
dissolved. Once dissolved, maintain in the dark at room tem-
perature. Right before use, warm the dissolved GelMA solution
at 37 C, add PI (final concentration 0.1%), and filter with a
0.22 μm vacuum filter system (see Note 2).
3. Trypsinize flasks of cultured dental epithelial (DE) and dental
mesenchymal (DM) cells (see Note 3). Collect, wash, and
count the cells, and resuspend DE and DM cell pellets each in
the appropriate filtered GelMA/PI solution (3% GelMA for
DE cells and 5% GelMA for DM cells) at 30–60 x 106 cells/
mL. Mix well (Fig. 1a1).
4. In the tissue culture hood, place individual 6 mm-inner-diam-
eter polydimethylsiloxane (PDMS) ring-shaped molds in the
center of each well of a 24-well tissue culture-treated plate
(Fig. 1a2) (see Note 4). Pipette 40–50 μL of GelMA cell
suspension into each PDMS mold. Place sample directly under

the UV spot curer at a distance of 6 cm, and photocrosslink by
UV exposure at 9.16 W/cm2 for 30–35 s (see Note 5)
(Fig. 1a3).
5. Leave photocrosslinked samples in molds for 5 min, and then
carefully remove the PDMS molds from each sample using
forceps (Fig. 1a4). Next, add 1 mL of BTB culture media to
each sample well. Culture cell-encapsulated GelMA constructs
in a humidified incubator with 5% CO2 at 37 C for 1–14 days.
Change media every 2–3 days.
3.3 Live/Dead 1. Wash GelMA tooth bud constructs in the culture plate 3 in
Analysis of In Vitro PBS to remove all media and serum.
Cultured GelMA Tooth 2. Prepare 1 μM calcein-AM and 4.5 μM ethidium homodimer
Bud Constructs (EthD-1) solution in PBS. Add 2 mL of calcein/ethidium
solution to each sample well, and incubate in a humidified
incubator with 5% CO2 at 37 C for 30 min.
3. Transfer stained samples from culture plate to a new plate
containing 2 mL of PBS per well.
4. Immediately image the stained samples using confocal
microscopy.
3.4 Subcutaneous 1. After 1 week in vitro culture in osteogenic media, wash

Implantation, Harvest, cultured tooth bud constructs 3 in 2 mL PBS. Carry washed
and Fixation of GelMA samples in covered 24-well plates to animal facilities.
Tooth Bud Constructs 2. Create four subcutaneous implantation pockets (two on each
side) on the back of each anesthetized rat host by making 1 cm
incisions about 1.5 cm away from the midline (Fig. 1b1). Place
one GelMA tooth bud construct in each subcutaneous pocket
(Fig. 1b2) (see Note 6). Close each incision with wound clips.
Remove wound clips after 2 weeks.
3. To harvest implanted tooth bud constructs, euthanize the rat
host and make a ~6 cm incision at the midline. Next, make
another incision at the top of and perpendicular to the midline
incision, also ~6 cm long. Make a similar incision at the bottom
of the midline incision. Use a scalpel and forceps to remove the
skin and subcutaneous layer from the back of the host
(Fig. 1b3).
4. Use a scalpel to cut a square around each implant. Use forceps
to peel the implant and subcutaneous tissue from the skin, and
place each sample in individual 5 mL sample collection bottle
washed 3 in PBS.
5. Immediately fix harvested samples in 10% formalin overnight at
room temperature.
6. Wash samples 3 in PBS and store in PBS at 4 C.
3.5 Processing and 1. Characterize mineralized tissue formation in fixed GelMA con-
Analyses of GelMA structs using X-ray or microCT. Begin decalcification of miner-
Tooth Bud Constructs alized constructs by immersing in 5 mL of decalcification
solution and gently rocking at room temperature. Change
3.5.1 Decalcification and decalcification solution every 24–48 h.
Processing of GelMA Tooth
Bud Constructs 2. Monitor decalcification every 24–48 h via ammonium oxalate-
calcium precipitation assay; remove 5 mL of harvested decalci-
fication solution, place in a small glass specimen bottle, and add
1 mL of saturated ammonium oxalate. Watch for precipitate
formation after 20 min at room temperature. Continue to
decalcify sample if precipitate forms. If precipitation does not
form within 20 min, wash the demineralized sample 3 in PBS
and store in fresh PBS at 4 C.
3. To process harvested and demineralized constructs, place each
tooth bud construct between two tissue processing sponges,
and secure within tissue cassettes labeled with pencil. Immedi-
ately immerse in graded ethanol (50, 70, 80, 90, and 100%) for
2–4 h each. Then immerse samples 2 in 100% ethanol for 2 h
each and then 3 in 100% xylenes 1–2 h each.
3.5.2 Paraffin Embedding 1. Immediately after processing, incubate tooth bud constructs in
and Sectioning molten paraffin for 12–16 h, 2. Once fully infiltrated with
paraffin, embed tooth bud constructs in molten paraffin using a
plastic mold, orienting with pre-warmed forceps (see Note 7).
2. Once embedded, place each construct on a cold plate or cryo
console (e.g., Microm EC 350-2, Thermo Fisher Scientific) for
1 h to solidify, and then allow to set at room temperature
overnight. Place at 4 C for long-term storage.
3. Use a microtome to section paraffin block containing samples
into 6 μm sections.
4. Float sections onto a microscope slide using a tissue floating
bath set at 45 C. Allow the slides to dry for at least 15 min at
room temperature. Place section-mounted slides on a hot plate
at 45 C for 1 h and then at 55 C overnight. Store mounted
sections in slide boxes at room temperature.
3.5.3 Hematoxylin and 1. Deparaffinize sections in 100% xylenes 2 for 5 min each.
Eosin Staining 2. Rehydrate deparaffinized sections in graded ethanol
(100, 95, 70, and 50%) 2 min each, and then rinse in tap
H2O for 2 min.
3. Stain sections with Mayor’s hematoxylin working solution for
1 min, followed by rinsing with tap water for 2.5 min.
4. Dip hematoxylin-stained sections once into diluted hydrochlo-
ric acid. Next, dip 3 into ammonium hydroxide working
solution.
5. Rinse sections in H2O for 2 min, and then place in eosin

working solution for 20 s.
6. Dehydrate H&E-stained sections by dipping 6 each in 4 dif-
ferent ethanol baths (95, 95, 100, and 100%).
7. Clear in 100% xylenes, 2 5 min each.
8. Mount coverslips over stained sections using Permount
mounting media, and let set overnight.
3.5.4 Immunohisto- 1. Deparaffinize sections in 100% xylenes 2 5 min.

chemical Staining Using 2. Rehydrate sections in graded ethanol (100, 95, 70, and 50%)
Vectastain Kit and DAB 2 min each and then in H2O for 5 min.
Detection
3. Block endogenous peroxidase activity by incubating sections in
3% H2O2 for 20 min at room temperature in the dark (e.g., in
an opaque slide container). Wash 3 in PBS for 10 min each.
4. Place slides in warmed TE buffer in a steamer for 20 min (see
Note 8). Cool slides in TE buffer on benchtop for 30 min.
Wash 3 in PBS for 5 min each.
5. Block sections with 10% donkey serum for 15 min at 37 C in a
humidified chamber (see Note 9).
6. Remove blocking solution, and incubate sections with primary
antibody diluted in 2% donkey serum for 1 h at room tempera-
ture in a humidified chamber (see Note 10). Then wash 3 in
PBS for 5 min each.
7. Incubate with secondary antibody diluted in 2% donkey serum
for 45 min at room temperature in a humidified chamber. Then
wash 3 in PBS for 5 min each.
8. Incubate sections in the dark with ABC reagent at room tem-
perature for 45 min in a humidified chamber. Then wash slides
3 in PBS for 5 min each.
9. In the light, incubate with DAB substrate at room temperature
for 5 min. Incubate in H2O for 5 min.
10. Counterstain sections in 0.2% Fast Green for 30–120 s (see
Note 11).
11. Next, dehydrate sections by dipping slides into graded ethanol
baths (95, 95, 100 and 100%) 3 each, and clear by dipping
3 in each 100% xylenes 2.
12. Mount coverslips over sections using Permount mounting
media, and let set overnight.
13. Image with Zeiss Axiophot Imager compound microscope
(Carl Zeiss, Zi) equipped with Axiophot digital camera (Carl
Zeiss AG, HRC) and AxioVision Rel 4.7 software (Fig. 1c).
3.5.5 Immunofluore- 1. Deparaffinize sections in 100% xylene 2 5 min each.

scence Staining 2. Rehydrate sections in a graded ethanol series (100, 95, 70, and
50%) 2 min each and then in H2O for 5 min.
3. Transfer slides to a plastic holder filled with pre-warmed
10 mM citric acid solution, and incubate in a steamer for
20 min (see Note 8). Cool at room temperature citric acid
solution on benchtop for 30 min. Wash 3 in PBS for
5 min each.
4. Place in a humidified chamber, and block with 5% BSA for
20 min at 37 C (see Note 9).
5. Remove blocking solution, and incubate sections with primary
antibody diluted with 1% BSA for 1 h at room temperature in a
humidified chamber (see Note 10). Wash slide 3 in PBS for
5 min each.
6. Incubate with secondary antibody diluted with 1% BSA for 1 h
at room temperature in a humidified chamber. Wash slides 3
in PBS for 5 min each.
7. Add approximately 10 μL of Hard Set Mounting Medium with
DAPI over each section, and mount with cover slips. Store at
4 C.
8. Image within 24 h with Zeiss Axiophot Imager compound
microscope (Carl Zeiss, Zi) equipped with Axiophot digital
camera (Carl Zeiss AG, MRM) and AxioVision Rel 4.7
software.
4 Notes
1. When calculating the final GelMA concentration and volume,

be sure to account for the final photoinitiator
(PI) concentration and volume that will be added at a later step.
2. Filter GelMA/PI solution using a Steriflip Vacuum Filtration
System (Millipore, SCGP00525).
3. Human umbilical vein endothelial cells (HUVECs) [9] or
other cell types can be co-encapsulated with the dental cells.
If incorporating other cells with the dental cells, be sure to
supplement the odontogenic media with cell type-specific
appropriate media. Also remember to include acellular control
GelMA control samples to monitor GelMA over time in culture
and for comparison with cell-encapsulated GelMA constructs.
4. Place a PDMS mold in every other well of a 24-well plate,
totaling 12 molds per plate. This minimizes UV exposure
from spot curing of adjacent samples.
5. The UV photocrosslinking parameters were optimized for the

OmniCure S2000 (Lumen Dynamics Group) equipped with a
5 mm spot curer and 320–500 nm filter and calibrated to
1.8 W.
6. For in vivo subcutaneous implantation of constructs, use a
randomization program to pre-assign samples to rat implanta-
tion pockets.
7. To limit the number of embedded sample blocks, include
replicate samples in a single paraffin block.
8. Refer to manufacturer’s antibody data sheet for specific condi-
tions for recommended antigen retrieval/unmasking and dilu-
tion ranges.
9. Prior to blocking, use a PAP pen to create a hydrophobic
barrier around each section to minimize the volume of block
solution and diluted antibody needed.
10. Include a no primary control to test for non-specific binding of
the secondary antibody. Add 2% donkey serum in place of the
primary antibody.
11. Refilter 0.2% Fast Green counterstain prior to use to remove
any precipitate.
Acknowledgments
All members of the Yelick Tissue Engineering Lab have contributed

to optimize and validate these techniques. This work was supported
by NIH/NIDCR R01 DE16132 (PCY) and NIH/NIDCR F31
DE026361 (EES).
References
1. Greenstein G, Cavallaro J, Romanos G, Tar- 5. Yen AH, Yelick PC (2011) Dental tissue regen-
now D (2008) Clinical recommendations for eration – a mini-review. Gerontology 57:85–94
avoiding and managing surgical complications 6. Smith EE, Yelick PC (2016) Progress in bioen-
associated with implant dentistry: a review. J gineered whole tooth research: from bench to
Periodontol 79:1317–1329 dental patient chair. Curr Oral Health Rep 3
2. Jung RE, Pjetursson BE, Glauser R, Zembic A, (4):302–308
Zwahlen M, Lang NP (2008) A systematic 7. Lai W-F, Lee J-M, Jung H-S (2014) Molecular
review of the 5-year survival and complication and engineering approaches to regenerate and
rates of implant-supported single crowns. Clin repair teeth in mammals. Cell Mol Life Sci
Oral Implants Res 19:119–130 71:1691–1701
3. Chrcanovic BR, Albrektsson T, Wennerberg A 8. Monteiro N, Smith EE, Angstadt S, Zhang W,
(2014) Reasons for failures of oral implants. J Khademhosseini A, Yelick PC (2016) Dental
Oral Rehabil 41:443–476 cell sheet biomimetic tooth bud model. Bio-
4. Chrcanovic BR, Kisch J, Albrektsson T, Wen- materials 106:167–179
nerberg A (2016) Factors influencing early 9. Smith EE, Zhang W, Schiele NR,
dental implant failures. J Dent Res Khademhosseini A, Kuo CK, Yelick PC
95:995–1002 (2017) Developing a biomimetic tooth bud
model. J Tissue Eng Regen Med 11 (2015) Synthesis, properties, and biomedical
(12):3326–3336 applications of gelatin methacryloyl (GelMA)
10. Monteiro N, Yelick PC (2016) Advances and hydrogels. Biomaterials 73:254–271
perspectives in tooth tissue engineering. J Tis- 12. Nichol JW, Koshy ST, Bae H, Hwang CM,
sue Eng Regen Med 11(9):2443–2461 Yamanlar S, Khademhosseini A (2010) Cell-
11. Yue K, Santiago GT-d, Alvarez MM, laden microengineered gelatin methacrylate
Tamayol A, Annabi N, Khademhosseini A hydrogels. Biomaterials 31:5536–5544
Chapter 15
Whole-Mount In Situ Hybridization of Mouse Embryos

Using DIG-Labeled RNA Probes
Abstract
Whole-mount in situ hybridization (WMISH) is a commonly used technique for visualizing the expression
profile of mRNAs in embryos. Unlike traditional in situ hybridization techniques, which require thin tissue
sections, the WMISH technique allows gene expression patterns to be assessed over the entire embryo and
structure. Here, we describe the detailed procedural steps of WMISH, including probe production, embryo
fixation and staining, and post-hybridization signal detection. Using this protocol, we visualized highly
specific expression patterns of Sonic hedgehog and Bmp4 mRNAs in E12.5 mouse embryos.
Key words Whole-mount, In situ hybridization, RNA probe, Digoxigenin labeling, Gene expression,
RNase-free, Mouse embryo, Craniofacial development
1 Introduction
Whole-mount in situ hybridization (WMISH) is a method of loca-

lizing and detecting specific RNA sequences in morphologically
preserved tissues by hybridizing the antisense probe to the RNA
sequence of interest. WMISH was first proposed in the late 1980s
as a modification of the traditional radioactive in situ hybridization
to study the gene expression pattern in Drosophila embryos [1]. In
the past decades, it has been widely used in developmental biology
for examining the gene expression in various species. In this pro-
cess, synthetically produced RNA probes are first complementarily
bound, or “hybridized,” to the transcripts of target genes. Immu-
nohistochemistry or fluorescence is then used to detect these RNA
hybrids, revealing spatial and temporal patterns of the gene expres-
sion. The advantage of WMISH is that it can provide an overall view
of a specific gene expression pattern on the whole embryo to
facilitate the observation of a gene of interest during embryo devel-
opment. The WMISH protocol we described here is composed of
151
152 Jingyi Wu and Xiaofang Wang
the following steps: (1) synthesis of DIG-labeled antisense probes,

(2) collection and fixation of embryos, (3) permeabilization of
embryos to facilitate probe penetration, (4) hybridization followed
by thorough washes, and (5) immunohistochemical detection of
the hybrids.
2 Materials
2.1 DNA Template 1. Sterilized PCR tubes and tips.

Preparation 2. PCR Master Mix.
3. Deionized H2O.
4. PCR thermal cycler.
5. RNeasy Mini Kit (Qiagen).
6. Reverse Transcription Kit (Qiagen).
7. QuickClean PCR Purification Kit (GenScript).
8. Agarose gel electrophoresis system.
9. NanoDrop.
10. SpeedVac concentrator.
2.2 RNA Probe 1. RNase-free PCR tubes and tips.

Preparation 2. 10 digoxigenin (DIG) labeling mixture (Roche).
3. RNA polymerase: T3, T7, or Sp6 (Roche).
4. RNase inhibitor.
5. Water bath.
6. Pre-cold 100% ethanol.
7. Pre-cold 70% ethanol in DEPC-H2O.
8. RNase-free DNase I.
9. 0.2 M EDTA (pH 8.0).
10. 4 M LiCl.
11. Refrigerated benchtop centrifuge.
12. 80 C freezer.
13. RNase-free water.
2.3 Embryo 1. Diethyl pyrocarbonate (DEPC).

Preparation 2. DEPC-PBS (see Note 1).
3. Dumont ultrafine tweezers.
4. 4% DEPC-paraformaldehyde (PFA).
5. Graded methanol series (25–100%) prepared with DEPC-
H2O.
Whole Mount ISH 153
Table 1
Composition of the hybridization buffer
Reagents Volume (mL)

Formamide 50
20 SSC (pH 4.5) 25
10%SDS 10
20 mg/mL tRNA 0.25
100 mg/mL heparin 0.05
DEPC-H2O Add to 100
6. RNase-free petri dishes.

7. RNase-free 24-well cell culture plates.
2.4 Hybridization 1. RNase-free proteinase K.

2. Glycine.
3. Hybridization buffer (Table 1).
4. Tris.
5. Sodium citrate.
6. Tween 20.
7. NaCl.
8. Blocking reagent (Roche).
9. Anti-digoxigenin antibody (Roche).
10. MgCl2.
11. BM purple (Sigma).
3 Methods
All procedures are performed at room temperature unless other-

wise indicated.
3.1 DNA Template 1. Extract total RNA from the head of E13.5 mouse embryos
Preparation using an RNeasy Mini Kit (Qiagen). Synthesize the first strand
cDNA using a Reverse Transcription Kit (Qiagen) following
the manufacturer’s instructions.
2. Design PCR primers for amplifying a specific segment
(~300–1000 bp length) of gene of interest. To facilitate the
future RNA probe transcription, the 50 end of reverse primer is
incorporated with an RNA polymerase promoter sequence.
Table 2
Composition of the transcription mix for synthesizing the DIG-labeled RNA
probes
Reagents Volume (μL)

Template cDNA 5 (1–4 μg)
10 Transcription buffer 2
10 Digoxigenin labeling mixture 2
RNase inhibitor 1
RNA polymerase (T3, T7, or Sp6) 2
RNase-free H2O Add to 20
3. PCR to amplify the cDNA segment of gene of interest (see

Note 2).
4. Purify the PCR products with a QuickClean PCR Purification
Kit (GenScript) following the manufacturer’s instructions.
5. Take 2–4 μL purified PCR products for agarose gel electropho-
resis to estimate the DNA amount by comparing to a standard
DNA ladder (or, the DNA amount can be determined by
NanoDrop analysis).
6. Concentrate the PCR products to make an appropriate volume
(5–10 μL) using a SpeedVac concentrator (see Note 3).
3.2 RNA Probe All procedures of the RNA probe synthesis must be RNase-free.
Synthesis
1. Prepare a 20 μL transcription mix (Table 2) for synthesizing the
DIG-labeled RNA probes.
2. Incubate the transcription mix in a water bath for 2 h at 37 C.
3. Add 2 μL RNase-free DNase I (20 U), and incubate at 37 C
15 min to remove the DNA template.
4. Add 2 μL 0.2 M EDTA (pH 8.0) to stop the reaction.
5. Add 2.5 μL 4 M LiCl and 75 μL pre-cold 100% ethanol to
precipitate the probe. Incubate at 80 C for at least 30 min.
6. 14,000 g, centrifuge for 15 min at 4 C.
7. Remove the supernatant, and wash the pellets with 50 μL
pre-cold 75% ethanol.
9. Dry the pellets and resuspend them in 50 μL DEPC-H2O.
10. Take 2 μL aliquot of the RNA probe for electrophoresis on a
1% agarose gel to estimate the concentration by comparing to a
standard DNA ladder (see Note 4). Alternatively, the concen-
tration of RNA probe can be measured by NanoDrop.
Whole Mount ISH 155
11. Add 2 μL RNase inhibitor to the probe, and split 10 μL ali-

quots for long-term storage. The RNase inhibitor will keep the
probe from degrading for several years at 80 C.
3.3 Mouse Embryo 1. Sacrifice the pregnant mice and dissect their uterus to expose
Collection and the embryos.
Preparation 2. Transfer the embryos to pre-cold DEPC-PBS in petri dishes.
Dissect the heads or lower jaws from the embryos under a
stereomicroscope with Dumont ultrafine tweezers (see Note
5).
3. Fix the embryos in a 24-well plate containing 4% PFA/DEPC-
PBS with gentle agitation at 4 C overnight.
4. Perform two consecutive washes with DEPC-PBST for
5 min each.
5. Dehydrate the embryos with graded methanol/DEPC-PBST
series (25, 50, and 75%) for 5 min each.
6. Bleach the embryos with 5% H2O2/methanol for 15–40 min at
room temperature (see Note 6).
7. Dehydrate the embryos with 100% methanol three times for
5 min each (see Note 7).
3.4 In Situ 1. Rehydrate the embryos with graded methanol/DEPC-H2O

Hybridization (Day 1) series (75, 50, and 25%) for 5 min each.
2. Wash with DEPC-PBST twice for 5 min each.
3. Permeabilize the embryos with 10 μg/mL RNase-free protein-
ase K at room temperature (see Note 8).
4. Wash with 2 mg/mL glycine/DEPC-PBST for 5 min at room
temperature.
5. Rinse the embryos with DEPC-PBST twice for 5 min each.
6. Postfix the embryos with 4% PFA for 20 min at room
temperature.
7. Wash with DEPC-PBST twice for 5 min each.
8. Add ~300 μL hybridization buffer to pre-hybridize the
embryos for 1 h at 70 C (see Note 9).
9. Denature the RNA probes (1 μg/mL in hybridization buffer)
for 15 min at 80 C and then immediately cool down on ice.
10. Add RNA probes to the embryos and incubate at 70 C
overnight.
3.5 Post- 1. Prepare washing buffers (Table 3) and preheat to 70 C.

hybridization Washes 2. Perform three consecutive washes with Buffer I for 30 min each
(Day 2) at 70 C (see Note 10).
3. Perform three consecutive washes with TBST (Table 4) for
5 min each at room temperature.
Table 3
Composition of the post-hybridization washing buffers
Reagents Buffer I (mL) Buffer II (mL)

Formamide 150 150
20 SSC (pH 4.5) 75 30
10% SDS 30 –
Deionized H2O Add to 300 Add to 300
Table 4
Composition of the TBST buffer
Reagents TBST (mL)

1 M Tris (pH 7.5) 50
5 M NaCl 15
Tween 20 0.5
Deionized H2O Add to 500
Table 5
Composition of the NTMT buffer
Reagents NTMT (mL)

2 M Tris (pH 9.5) 15
5 M NaCl 6
1 M MgCl2 15
Tween 20 0.3
3.6 Immunological 1. Incubate the embryos in 2% blocking reagent for 1–2 h at room
Incubation (Day 2) temperature.
2. Replace the blocking reagent with anti-digoxigenin antibody/
blocking solution (1:5000), and incubate overnight at 4 C.
3.7 Washes and 1. Bring the 24-well plate from 4 C to room temperature.
Color Development 2. Perform six consecutive washes with TBST for 60 min each to
(Day 3) remove any unbound antibody.
3. Perform three consecutive washes with NTMT (Table 5) for
Whole Mount ISH 157
Fig. 1 The whole-mount ISH shows Bmp4 (left) and Shh (right) expression patterns in the craniofacial tissues
of the E12.5 mouse embryos
4. Immerse the embryos in BM purple AP substrate at room

temperature in the dark to allow color development. Examine
the signal periodically under a stereomicroscope, and stop the
reaction by washing with PBST when the signal is suitable.
5. Postfix the slides with 4% PFA for 20 min at room temperature.
6. Rinse with PBST twice for 5 min each.
7. Take pictures using a stereomicroscopic system (Fig. 1).
8. The embryos can be stored in PBS or glycerol for several
months at 4 C.
4 Notes
1. All glassware and metal containers should be baked at 180 C

overnight. Plasticware should be treated with DEPC or RNase
ZAP and rinsed with RNase-free water before use. The solu-
tions used for in situ hybridization on Day 1 should be RNase-
free by adding 0.1% DEPC (v/v) and stirred overnight, fol-
lowed by autoclaving to remove the DEPC. Note that PFA
cannot be autoclaved and should be prepared with DEPC-PBS.
2. The PCR products purified from 2 to 4 50 μL volume reactions
should be sufficient for RNA probe transcription.
3. The purified DNA should be no less than 3 μg.
Table 6
Recommended duration of the proteinase K digestion according with the
embryonic stage of the specimens
Embryonic stage Incubation time with PK (min)

E9.5 5
E10.5 8
E11.5 12
E12.5 15
Table 7
Recommended volume of the hybridization buffer according to the
Embryonic stage Buffer volume (μL)

E9.5–10.5 300
E11.5–12.5 500
E13.5–14.5 700
>E15.5 1000
4. The electrophoresis of RNA probe should be no longer than

30 min to avoid degradation.
5. Remove unnecessary tissues as much as possible to fully expose
the desired region of interest.
6. The H2O2 solution should be freshly made. The bleaching
time of H2O2 depends on the size and developmental stage
of the embryos.
7. The embryos can be stored in 100% methanol for 1–2 months
at 20 C.
8. The incubation time of proteinase K depends on the develop-
mental stage and size of the embryos (Table 6). Note that over-
digestion will damage the embryo morphology.
9. The volume of hybridization buffer depends on the size and
developmental stage of the embryos (Table 7).
10. The embryos incubated at 70 C overnight will be sticky and
fragile. Caution should be taken in replacing the buffer to
avoid damaging the samples.
11. The duration of color development may vary between genes. It
may take only 15–30 min for some highly abundant transcripts
to show a satisfied signal, while for most other genes, the color
reaction generally lasts for 1–2 h. Some low-copy transcripts
may need several days to fully develop the color reaction.
Whole Mount ISH 159
References
1. Tautz D, Pfeifle C (1989) A non-radioactive in translational control of the segmentation gene
situ hybridization method for the localization of hunchback. Chromosoma 98:81–85
specific RNAs in Drosophila embryos reveals
Part III
Protocols for Studying Gene and Protein Expression

Chapter 16
In Situ Hybridization on Mouse Paraffin Sections

Using DIG-Labeled RNA Probes
Jingyi Wu, Jian Q. Feng, and Xiaofang Wang
Abstract
In situ hybridization is a commonly used technique using an antisense RNA probe to localize a specific RNA
sequence on histological sections. This approach can visualize the expression pattern of a gene of interest in
a portion of tissues. Here, we detail an optimized method for performing in situ hybridization on mouse
paraffin sections using digoxigenin (DIG)-labeled RNA probes.
Key words In situ hybridization, RNA probe, Digoxigenin labeling, Gene expression, Paraffin
section, RNase-free, mRNA
1 Introduction
The principle behind in situ hybridization (ISH) is the specific

annealing of a labeled nucleic acid probe to complementary
sequences in fixed tissue, followed by the visualization of the
probe location. ISH permits the precise cellular localization and
identification of cells expressing a particular gene during craniofa-
cial development [1, 2]. In this protocol, we focus on detecting
gene expression on the paraffin sections of mouse dental tissues
with ISH using DIG-labeled antisense RNA probes. This protocol
describe the detailed steps of ISH, including probe production,
embryo collection and section preparation, and hybridization and
post-hybridization treatments. Using this protocol, we visualized
highly specific expressions of Sonic hedgehog and ameloblastin
mRNAs in the lower incisors of E13.5 and E18.5 mouse embryos.
2 Materials
2.1 DNA Template 1. Sterilized PCR tubes and tips.

Preparation 2. Plasmid containing the cDNA of the gene of interest.
163
164 Jingyi Wu et al.
3. Restriction enzymes and buffers (NEB).

4. Water bath.
5. QuickClean PCR Purification Kit (GenScript).
6. Agarose gel electrophoresis system.
7. NanoDrop.
8. SpeedVac concentrator.
9. Deionized H2O.
2.2 RNA Probe 1. RNase-free PCR tubes and tips.

Preparation 2. 10 digoxigenin labeling mixture (Roche).
3. RNA polymerase: T3, T7, or Sp6 (Roche).
4. 10 transcription buffer (Roche).
5. RNase inhibitor.
6. Water bath.
7. Pre-cold 100% ethanol.
8. Pre-cold 70% ethanol in DEPC-H2O.
9. RNase-free DNase I.
10. 0.2 M EDTA (pH 8.0).
11. 4 M LiCl.
13. 80 C freezer.
14. RNase-free water.
2.3 Embryo 1. Diethyl pyrocarbonate (DEPC).

Collection and Slide 2. 200 C oven (see Note 1).
Preparation
3. DEPC-PBS.
4. Dumont ultrafine tweezers.
5. RNase-free Petri dishes.
6. 4% DEPC-paraformaldehyde (PFA) (see Note 2).
7. 50–100% gradient ethanol in DEPC-H2O.
8. Xylene.
2.4 Hybridization 1. RNase-free proteinase K.

2. Glycine.
3. Hybridization buffer (Table 1).
4. Tris base.
5. Sodium citrate.
6. Tween 20.
7. NaCl.
ISH on Tissue Sections 165
Table 1
Composition of the hybridization buffer
Reagents Volume (mL)

Formamide 50
20 SSC (pH 4.5) 25
10% SDS 10
tRNA (20 mg/mL) 0.25
Heparin (100 mg/mL) 0.05
DEPC-H2O Add to 100 mL
8. Blocking reagent (Roche).

9. Anti-digoxigenin antibody.
10. MgCl2.
11. BM purple (Sigma).
12. Mounting medium.
3 Method
All procedures are performed at room temperature unless other-

wise indicated.
3.1 DNA Template 1. Linearize ~4 μg plasmid containing the cDNA of the gene of
Preparation interest with a suitable restriction enzyme in a 20 μL system in a
water bath following the manufacturer’s instructions.
2. When digestion is completed, take 2 μL of the linearized DNA
for electrophoresis on a 1.5% agarose gel. The complete linear-
ization of plasmid DNA will show a single band on the gel at
the designated position (see Note 3). The linearized DNA can
be stored at 20 C for several weeks.
3. Purify the linearized DNA using a QuickClean PCR Purifica-
tion Kit according to the manufacturer’s instructions.
4. Concentrate the DNA to an appropriate volume (5–10 μL)
using a SpeedVac concentrator.
3.2 RNA Probe All procedures of the RNA probe synthesis must be RNase-free.
Preparation
1. Prepare a 20 μL transcription mix for synthesizing the
DIG-labeled RNA probes (Table 2).
2. Incubate the transcription mix in a 37 C water bath for 2 h.
3. Add 2 μL RNase-free DNase I (20 U), and incubate at 37 C
for 15 min to remove the DNA template.
Table 2
Composition of the transcription mix for synthesizing the DIG-labeled RNA
probes
Reagents Volume
Template DNA 5 μL (1–4 μg)
10 Transcription buffer 2 μL
10 Digoxigenin labeling mixture 2 μL
RNase inhibitor 1 μL
RNA polymerase (T3, T7, or Sp6) 2 μL
RNase-free H2O Add to 20 μL
4. Add 2 μL 0.2 M EDTA (pH 8.0) to stop the reaction.

5. Add 2.5 μL 4 M LiCl and 75 μL pre-cold 100% ethanol to the
tube and mix well. Incubate at 80 C for at least 30 min.
7. Remove the supernatant and wash the pellets with 50 μL
pre-cold 75% ethanol.
9. Dry the pellets at room temperature and resuspend in 50 μL
DEPC-H2O.
10. Take 2 μL aliquot of the RNA probe for electrophoresis on a
1% agarose gel to estimate the concentration by comparing to a
standard DNA ladder (see Note 4). Alternatively, the concen-
tration of RNA probe can be measured by NanoDrop.
11. Add 2 μL RNase inhibitor to the probe, and split 10 μL ali-
quots for longer-term storage. The RNase inhibitor will keep
the probe from degrading for several years at 80 C.
3.3 Embryo 1. Sacrifice the pregnant mice and dissect their uterus to expose
Collection and Slide the embryos.
Preparation 2. Transfer the embryos to pre-cold DEPC-PBS in Petri dishes.
Dissect the heads or lower jaws from the embryos under a
stereomicroscope with Dumont ultrafine tweezers (see
Note 5).
3. Fix the embryos in a clean 50 mL centrifuge tube containing
4% PFA in DEPC-PBS at 4 C with gentle rocking overnight.
4. Wash the embryos with DEPC-PBS twice for 5 min each.
Dehydrate the embryos with graded ethanol/DEPC-H2O
series (50, 70, 80, 90, and 100%).
Table 3
The series of xylene and graded ethanol treatments for the deparaffination
and rehydration of the slides
Reagents Time (min)

Xylene 10
Xylene 10
100% ethanol 5
100% ethanol 3
90% ethanol 2
70% ethanol 2
50% ethanol 2
5. Immerse the embryos in 1:1 ethanol/xylene for 30 min and

then 100% xylene to make transparent the embryos (see
Note 6).
6. Perform paraffin infiltration overnight and then embedding.
7. Cut 4–7-μm-thick sections and dry the slides at 37 C over-
night (see Note 7).
3.4 Pre-hybridization 1. Deparaffin and rehydrate the slides through a series of xylene
Treatment (Day 1) and graded ethanol treatments (Table 3).
2. Perform two consecutive washes with DEPC-PBST (1:1000
Tween 20/DEPC-PBS) for 5 min each at room temperature.
3. Permeabilize the slides in 10–20 μg/mL proteinase K/DEPC-
PBST with gentle agitation at room temperature. The protein-
ase K concentration and digestion time are adjustable based on
the section thickness and embryonic stage of the sample (see
Note 8).
4. Wash in 2 mg/mL glycine/DEPC-PBST for 5 min.
5. Perform two consecutive washes with DEPC-PBST for
5 min each.
6. Postfix the digested slides for 20 min in 4% PFA in DEPC-PBS.
7. Perform three consecutive washes with DEPC-PBS for
5 min each.
8. At this point, denature the RNA probe (1 μg/mL in hybridiza-
tion buffer) for 15 min at 80 C and then immediately cool
down on ice (see Note 9). Preheat the hybridization chamber
to 65 C.
9. Rinse the slides with 2 DEPC-SSC (pH 4.5) for 5 min.
Table 4
Composition of the post-hybridization washing buffers
Reagents Buffer I (mL) Buffer II (mL)

Formamide 150 150
20 SSC (pH 4.5) 75 30
10% SDS 30 –
Deionized H2O Add to 300 Add to 300
Table 5
Composition of the TBST buffer
Reagents TBST (mL)

1 M Tris (pH 7.5) 50
5 M NaCl 15
Tween 20 0.5
3.5 Hybridization 1. Place slides on the racks of a preheated hybridization chamber.

(Day 1) Add 300 μL probe/hybridization buffer to each slide to fully
cover the sections.
2. Cover the slides with RNase-free coverslips (see Note 10).
Avoid any bubbles under the coverslips.
3. Hybridize overnight at 65 C (see Note 11).
3.6 Post- 1. Prepare washing buffers (Table 4) and preheat to 65 C.

hybridization Washes 2. Stop hybridization and gently remove the coverslips from the
(Day 2) slides.
3. Perform three consecutive washes with Buffer I for 30 min each
at 65 C.
4. Perform three consecutive washes with Buffer II for 30 min
each at 60 C (see Note 12).
5. Perform three consecutive washes with TBST (Table 5) for
5 min each at room temperature.
3.7 Immunological 1. Put the slides in a humid chamber, and incubate with 2%
Incubation (Day 2) blocking reagent for 1–2 h (see Note 13).
2. Replace the blocking reagent with anti-digoxigenin antibody/
blocking solution (1:2000), and incubate overnight at 4 C.
Table 6
Composition of the NTMT buffer
Reagents NTMT (mL)

2 M Tris (pH 9.5) 15
5 M NaCl 6
1 M MgCl2 15
Tween 20 0.3
Fig. 1 Left panel shows Shh expression in the dental epithelium of E12.5 mouse lower incisors. Right panel
shows ameloblastin expression in the ameloblasts of E18.5 mouse lower incisors. The red lines plot the border
between the dental epithelium and mesenchyme and the outline of the lower incisors. Unpublished data by
Dr. Jingyi Wu
3.8 Washing and 1. Bring the humid chamber from 4 C to room temperature.
Color Development 2. Perform six consecutive washes with TBST for 30 min each to
(Day3) remove any unbound antibody.
3. Perform three consecutive washes with NTMT (Table 6) for
4. Add 300 μL BM purple AP substrate (Roche) to each slide.
Cover the slides with parafilm, and incubate at room tempera-
ture in the dark to allow color development. Examine the signal
periodically under a microscope, and stop the reaction by
immersing the slides into PBST when the signal is suitable
(see Note 15) (Fig. 1).
5. Postfix the slides with 4% PFA for 20 min at room temperature.
6. Dehydrate the slides through a graded ethanol series (70, 95,
and 100%).
7. Immerse the slides in xylene for 3 min and mount in a suitable
organic mounting medium.
4 Notes
1. All glassware and metalware should be baked at 180 C over-

night. Plasticware should be treated with DEPC or RNase ZAP
and rinsed with RNase-free water before use. The solutions
used for in situ hybridization on Day 1 should be RNase-free
by adding 0.1% DEPC (v/v) and stirred overnight, followed by
autoclaving to remove the DEPC.
2. The PFA cannot be autoclaved and should be prepared with
DEPC-PBS.
3. Use the PCR Purification Kit to purify the linearized plasmid
DNA following the manufacturer’s instructions.
4. The electrophoresis of the RNA probe should be less than
30 min to avoid degradation of the probe.
5. The dissection should be finished in 3 min to avoid significant
degradation of the tissue RNAs.
6. The transparent procedure must be closely monitored as over-
treatment by xylene may cause embedding defects. Xylene
treatment must be stopped immediately once the embryos
get clear.
7. The slides can be stored in an RNase-free container at 20 C
for several months.
8. The time required for proteinase K (PK) digestion may vary in
the sections from different stage embryos (Table 7).
9. The volume of the hybridization buffer varies (100–300 μL),
depending on the number of samples on each slide.
10. The coverslips should be baked at 180 C overnight to remove
the RNase before use.
11. The hybridization temperature needs to be optimized with
each probe. We usually use 65 C for general hybridization.
12. The second wash temperature is usually 5 C lower than the
hybridization temperature.
Table 7
Recommended duration of the proteinase K digestion according to the
Embryonic stage Incubation time with PK (min)

E9.5 5
E10.5 8
E11.5 12
E12.5 15
13. The blocking reagent needs to be dissolved in advance. It may

take 0.5–1 h to fully dissolve the blocking reagent at 60 C.
14. Add levamisole (0.5 mg/mL) to remove endogenous alkaline
phosphatase in older (>E10.5) embryo samples.
15. The color reaction may take tens of minutes up to several days.
Stop the reaction when the signal is satisfied with a minimal
background.
References
1. Zhang Z, Song Y, Zhao X et al (2002) Rescue of 2. Wang X, Hao J, Xie Y et al (2010) Expression of
cleft palate in Msx1-deficient mice by transgenic FAM20C in the osteogenesis and odontogenesis
Bmp4 reveals a network of BMP and Shh signal- of mouse. J Histochem Cytochem 58:957–967
ing in the regulation of mammalian palatogen-
esis. Development 129:4135–4146
Chapter 17
Methods for In Situ Protein Visualization in Dental

Mineralized Tissues
D. Hotton, A. Berdal, and A. Bolaños
Abstract
Immunohistochemistry (IHC) is a technique based on the specificity of antibody-antigen principle used
commonly to detect antigens in tissue sections. The immune labeling can be performed in paraffin sections,
cryostat sections, and ultrathin sections and can be observed in light confocal and transmission electron
microscopy. However, the use of immunohistochemical techniques for the study of mineralized tissues has
been a challenge for decades (Berdal et al., Arch Oral Biol 36:715–725, 1991; Nanci et al., Eur J Histochem
52:201–214, 2008). Specific procedures are necessary when compared with soft tissue immunohistochem-
istry. This chapter describes methods for IHC on Tissue-Tek O.C.T. compound and paraffin-embedded
sections to detect antigens in the dental mineralized tissues.
Keywords Dental mineralized tissues, Cryostat sections, Paraffin sections, Immunoperoxidase,

Immunofluorescence
1 Introduction
The principle of IHC has existed since the 1930s, but it was not
until 1942 that the first IHC study was reported by Coons et al.
[1]. They used fluorescein isothiocyanate (FITC)-labeled
antibodies to localize pneumococcal antigens in infected tissues.
Developed from the antigen-antibody binding reaction, immuno-
histochemistry can be considered as a method that visualizes distri-
bution and localization of specific antigen or cellular components in
the tissue sections. Compared to other techniques that are based on
the antigen-antibody reaction such as immunoprecipitation or
western blot, immunohistochemistry provides in situ protein visu-
alization. The development and improvement of protein conjuga-
tion have allowed the use of enzyme markers mainly peroxidase and
alkaline phosphatase in light microscopy [2] and colloidal gold in
transmission electron microscopy (TEM) [3]. In addition, labeling
using fluorescent, radioactive, and chemiluminescent markers has
been frequently used [4]. Therefore, immunohistochemistry
173
174 D. Hotton et al.
(IHC) has become an essential and routine protocol in research

laboratories. This is a wide-used biological technique that combines
anatomy, physiology, immunology, and biochemistry. In general,
the protocols have been adapted and optimized for each tissue type
and primary antibody [5].
2 Materials
1. Phosphate-buffered saline (PBS): 10 solution (Euromedex,

France) and dilute 1 with distilled water; adjust pH to 7.4.
2. Carnoy fixation. His composition is chloroform 1/5, alcohol
1/5, and acetic acid 3/5 of solution.
3. Fixation solution, 4% paraformaldehyde (PFA) in 1 PBS: To
extemporaneous preparation of 400 mL of 4% PFA solution,
add 50 mL of 32% paraformaldehyde stock solution (Electron
Microscopy Science, LFG, France) (see Note 1) to 350 mL of
distilled water in a glass beaker. Add drops of 2N NaOH to
adjust the pH to 7.4.
4. Decalcifying solution 4.13 % EDTA (Sigma-Aldrich, France) in
distilled water. Dissolve 41.30 g EDTA: the solution becomes
cloudy. Add 40 tablets NaOH to adjust pH to 7.4 for 1 L (see
Note 2).
5. 30% sucrose (Merck, VWR, France) in 1 PBS at 4 C. Mix
30g sucrose in 1 L 1 PBS (see Note 3).
6. Tissue-Tek O.C.T. compound (Cryoblock, Labonord, France)
is a formulation of water-soluble glycols and resins, providing a
convenient specimen matrix for cryostat sectioning at tempera-
tures of 10 C and below.
7. SuperFrost® Plus slides (VWR, France), 25 mm 75 mm (see
Note 4).
8. Proteinase K treatment (see Note 5). Proteinase K buffer:
Tris–HCl 50 mM with 10 mM CaCl2 pH 6. Tris–HCl 1 M:
Dissolve 121.14 g Tris (VWR, France) in 800 mL distilled
water. Adjust pH to 8.0 with the appropriate volume of 1 M
HCl. The final volume is 1 L distilled water. CaCl21M: Dis-
solve 11,098 g CaCl2 in 800 mL distilled water. Add distilled
water to a final volume to 1 L; mix 50 mL Tris–HCl 1 M and
10 mL CaCl21M. Bring final volume to 1 L distilled water, and
adjust pH to 8.0. Autoclave and store at room temperature.
Proteinase K concentration stock solution: 20 mg/mL in dis-
tilled water at 20 C. Proteinase K concentration solution:
50 μg/mL.
9. 3% H2O2 in methanol (VWR, France). Mix 6 mL 30% H2O2 in
200 mL methanol (see Note 6).
Proteins in Dental Mineralized Tissues 175
10. Blocking solution: Background staining was reduced with 1

PBS/0.5% bovine serum albumin (BSA) (Life Science, Sigma-
Aldrich, France) and 0.1% Tween (Life Science, Sigma-Aldrich,
France) including 5% nonimmune serum from the same species
in which the secondary antibody is made in primary antibody
dilution solutions and secondary antibody dilution solution.
Make 10 mL with 0.05 g BSA, 10 μL Tween, and 0.5 mL
nonimmune serum.
11. Washing solution: 1 PBS/0.5% bovine serum albumin (BSA)
with 0.1% Tween.
12. Humidified chamber: Microscope slide staining dishes, racks,
and jars (Ted Pella, Inc.); 360 246 StainTray slide staining
system for 20 slides (VWR, France).
13. Goat anti-rabbit secondary antibody (Dako, Agilent, USA) (see
Note 7).
14. Goat anti-rabbit biotinylated secondary antibody and strepta-
vidin peroxidase (Vectastain® ABC kit, Vector, Clinisciences,
Paris) (see Note 8).
15. Protein A peroxidase (Molecular Probes, Thermo Fisher,
France) (see Note 9).
16. DAB solution ready to use is the peroxidase substrate solution
(ImmPACT NovaRED, Vector, Clinisciences, France).
17. Goat polyclonal secondary antibody to rabbit IgG Alexa
Fluor® 647 (Abcam, France) (see Note 10).
18. Hematoxylin (VWR, France).
19. DPX mounting medium (Life Science, Sigma-Aldrich, France).
20. 40 ,6-Diamidino-2-phenylindole (DAPI) Fluoromount-G® is a
water-soluble, containing compound recommended for slides
mounted after a staining procedure having an aqueous final
step (Southernbiotech, Clinisciences, France).
Liquid blocker Dako Pen (Dako Agilent Pathology Solutions,
USA). Figure 1 illustrates a confocal microscopy visualization of
collagen 1 in dental tissues from 3-month-old mice comparatively
to controls performed following the protocol described below (see
Fig. 1).
3 Methods
3.1 Tissue Dental tissue sample processing with two different fixation solu-
Processing tions and four visualization techniques used will be detailed as
follows:
Fig. 1 Confocal microscopy visualization of collagen 1 in dental tissues from 3-month-old mice. (a) Control
gingiva section of FAM20A mutant null mice without primary antibody. (b) Gingiva section of FAM20A mutant
null 3-month-old mice incubated with rabbit polyclonal collagen type 1 antibody (Abcam 34,710, 1/100
diluted) and goat anti-rabbit Alexa Fluor® 647 (Abcam 150,079, 1/400 diluted) and mounted with DAPI. Arrows
showed ectopic calcifications (EC). (c) Wild-type mice control root section without primary antibody. (d) Wild-
type mice first molar root section incubated with rabbit polyclonal collagen type 1 antibody (Abcam 34,710,
1/100 diluted) and goat anti-rabbit Alexa Fluor® 647 (Abcam 150,079, 1/400 diluted) mounted with DAPI. Pink
color showed collagen 1 expression in odontoblasts (OD) and osteoblasts (OS). The bone’s background was
detected with pale red color (*)
3.1.1 Cryostat Sections Carnoy fixation is used. After 16 h immersion fixation at 4 C and
washes in 1 PBS, we infuse the samples in 30% sucrose in 1 PBS
at 4 C (up to 1 day under gentle agitation) to reduce the risk of
damages during freezing with cryoblock. Sections with the cryostat
–25 C (CM 3050S, Leica, Rueil-Malmaison, France) are made
7 μm-thick and mounted on glass microscope SuperFrost® Plus
slides and kept overnight at 4 C. Allow slides to dry at room
temperature for 30 min. Wash for 5 min with 1 PBS.
3.1.2 Paraffin Sections Mice were anaesthetized and fixed with 4% paraformaldehyde in 1
PBS by a 15 min intracardiac perfusion through the left ventricle
using a monostaltic pump (Touzart & Matignon, Vitry, France)
followed by immersion in 4% paraformaldehyde in 1 PBS for 16 h
at 4 C. The mandibles were dissected. We rinsed with 1 PBS and

decalcified for 4 weeks at 4 C in 4.13% disodium ethylenediami-
netetraacetic acid (see Note 2), dehydrated, and embedded in
paraffin. Seven micron sections were cut and kept overnight at
40 C. After deparaffination, rehydrating sections were removed,
rehydrated, and rinsed 3 min in 1 PBS.
3.2 Protocols 1. Remove the wax from the paraffin tissue sections with Clearene
(Leica, France) 5 min twice.
2. Rehydrate the sections with ethanol 100% twice for 3 min,
ethanol 80% 3 min, and ethanol 70% 3 min.
3. Wash in phosphate buffer saline (PBS) 1 in distilled water.
4. 5 min in proteinase K buffer.
5. Proteinase K treatment was realized (5 min in PK buffer 50 μg/
mL at 37 C).
6. Rinse 2 for 5 min in washing solution.
7. Encircle the tissue sections with liquid blocker Dako Pen and
let dry for 10 s.
8. Incubate in blocking solution for 2 h at RT (room tempera-
ture) or 16 h at 4 C.
9. Incubate in blocking solution for 2 h at RT or 16 h at 4 C.
10. Rinsing not required in the next step.
11. Make up the appropriate dilution for the antibody in blocking
solution (see Note 11).
12. Incubate with primary antibody diluted in blocking solution at
RT for 1 h or at 4 C overnight in humidified chamber. (The
blocking solution alone may be used as a primary antibody
negative control.)
14. Quench endogenous peroxidase with 3% H2O2 in methanol
for 10 min.
16. Using the Dako Pen, a water repelling circle around tissue
sections can be drawn. A circle provides a barrier to liquids
such as antibody solutions applied to the sections, thus helping
to obtain more uniform immunohistochemical staining results
and reduce the amount of reagents.
17. Peroxidase revelation systems.
18. Incubate for 60 min at RT with diluted 1/400 HRP secondary
in blocking solution.
19. Or incubate for 30 min at RT with diluted biotinylated sec-
ondary antibody, and after rinsing, incubate for 30 min with
streptavidin peroxidase.
20. Or incubate for 60 min at RT in peroxidase protein A diluted

1/100 in blocking solution (see Note 9).
21. Rinse 2 for 5 min in washing solution, and incubate the
sections with ImmPACT NovaRED peroxidase substrate solu-
tion at room temperature for 2–15 min. Check the signal
development under a light microscope.
22. After peroxidase color development and optional counter col-
oration (hematoxylin), wash the slides in distilled water, and
dehydrate by passing through a series of alcohols (50, 75,
95, and 100%) and mount using DPX.
23. Fluorescent Alexa revelation system: Rinse 3 for 5 min in
washing solution, and incubate in goat polyclonal secondary
antibody to rabbit IgG Alexa Fluor® 647 diluted 1/400 (see
Note 10).
24. Wash 3 for 5 min in distilled water, and mount with DAPI
Fluoromount-G®.
4 Notes
1. Paraformaldehyde solution must be prepared extemporane-

ously to avoid the acid formic formation. PFA is extremely
toxic that should be diluted in a fume hood with rubber gloves
and a face mask as a minimum standard.
2. Complete decalcification of murine mandibles in 4.13% ethy-
lenediaminetetraacetic acid (EDTA) generally requires 1 to
6 weeks for 11- to 120-day-old mice, respectively. To accelerate
the decalcification for histological analysis, we now routinely
use a microwave oven (PELCO Bio Wave Pro®, Ted Pella Inc.).
Microwave irradiation offers a means of delivering energy
directly to the specimen while providing control over the
amount of heat generated (30 C) [6]. Several studies [7–9]
have confirmed the efficiency of this method (good preserva-
tion of the bone and dental tissues after microwave
decalcification).
3. For embedding the tissues before section, we used Tissue-Tek
O.C.T. compounds. The 30% sucrose solution in 1 PBS at
4 C (up to 1 day) was the cryoprotection for the tissues before
embedding to reduce the risk of tissue damages (development
of ice crystals) during freezing.
4. For long time, we handled glass slides with some gelatin, silane,
or the polylysine solutions for recovery. The arrival of Super-
Frost® Plus slides eliminates the need for these special adhesives
and protein coatings. SuperFrost Plus© slides are recommended
for hard tissue samples like bone. They are ideal for immuno-
cytochemical in dental mineralized tissues.
5. The proteinase K treatment step is critical for mineralized

tissues. Proteinase K is used for the proteolytic digestion of
paraffin-embedded, formalin-fixed mineralized tissues for trea-
ted crossed reactions with all cellular components. The pro-
teinase K treatment increases the permeability and thereby the
protocol efficiency. Insufficient digestion will reduce the signal,
and over-digestion will result in poor tissue morphology. The
extent of proteinase K treatment is determined with the pro-
teinase K titration experiment concentration.
6. Methanol, PBS, distilled water, or saline can be used to dilute
hydrogen peroxide. Morphology peroxidase-rich tissues could
be damaged by the aqueous hydrogen peroxide solution.
Therefore, methanol is a better choice in this case. Some cell
surface markers are very sensitive to methanol/hydrogen per-
oxide quenching, reducing the staining of antigenic site, par-
ticularly on frozen sections. So using hydrogen peroxide in PBS
is recommended for cell surface or membrane markers. We
don’t use antigen retrieval methods because high temperature
can cause damage to sections.
7. Goat anti-rabbit secondary antibody. This is the purified immu-
noglobulin fraction of rabbit antiserum conjugated with horse-
radish peroxidase (HRP) of very high specific enzymatic
activity (Dako, Agilent, USA). This classic indirect method
involves that the unlabeled primary antibody that reacts with
tissue antigen then reacts with a labeled secondary antibody.
This method is more sensitive due to signal amplification
through several secondary antibody reactions with different
antigenic sites on the primary antibody. The second antibody
can be labeled with a fluorescent dye such as Alexa and may be
labeled with an enzyme such as peroxidase.
8. Streptavidin-biotin complex (ABC) method: ABC method is a
standard IHC method that is a widely used technique for
immunohistochemical staining. Streptavidin is derived from
Streptococcus. Since the streptavidin molecule is uncharged,
the tissue electrostatic binding is eliminated. It can be labeled
with peroxidase or Alexa and has a very high affinity for biotin.
Biotin, a low molecular weight vitamin, can be conjugated to a
variety of biological molecules such as antibodies.
9. Protein A is a staphylococcal bacterial protein that binds with
high affinity to the Fc portion of various classes and subclasses
of immunoglobulins from a number of species (human, rabbit,
and goat).
10. Alexa Fluor® 647 conjugates exhibit more intense fluorescence
and are more photostable than most other fluorescent conju-
gates. A significant advantage of using this Alexa Fluor® 647 is
the low autofluorescence of biological specimens in this region
of the spectrum. It is most commonly visualized with a confo-

cal microscope equipped with an appropriate laser for excita-
tion and a far-red detector.
11. The optimal antibody concentration, which gives the best
staining with minimum background, must be determined
experimentally for each assay and is usually determined by
using a series of dilutions. The antibody dilutions span the
optimal concentrations between the dilutions recommended
by the supplier. In the case the dilution recommended is 1:200,
we utilized 1:100–1:200–1:400 ratio dilutions. This should
determine the optimal and most economical dilution. The
volume required will depend on the section size; we utilized
50 μL per section.
References
1. Coons AH, Creech HJ, Jones RN, Berliner E Bone research protocols, methods in molecular
(1942) The demonstration of pneumococcal biology, vol 816, 2nd edn. Springer, Heidelberg,
antigen in tissues by the use of fluorescent anti- pp 321–334
body. J Immunol 45:159–170 6. Cunningham CD, Schulte BA, Bianchi LM,
2. Jacques J, Hotton D, De la Dure-Molla M, Weber PC, Schmiedt BN (2001) Microwave
Petit S, Asselin A, Kulkarni AB, Gibson CW, decalcification of human temporal bones. Laryn-
Brookes SJ, Berdal A, Isaac J (2014) Tracking goscope 111(2):278–282
endogenous amelogenin and ameloblastin 7. Hellström S, Nilsson M (1992) The microwave
in vivo. PLoS One 16(6):9 oven in temporal bone research. Acta Otolaryn-
3. Papagerakis P, MacDougall M, Hotton D, gol Suppl 493:15–18
Bailleul-Forestier I, Oboeuf M, Berdal A 8. Keithley EM, Truong T, Chandronait B, Billings
(2003) Expression of amelogenin in odonto- PB (2000) Immunohistochemistry and micro-
blasts. Bone 32(3):228–240 wave decalcification of human temporal bones.
4. Ford PJ (2010) Immunological techniques: Hear Res 148(1–2):192–196
ELISA, flow cytometry and immunohistochem- 9. Katoh K (2016) Microwave-assisted tissue prep-
istry. In: Oral biology, methods in molecular aration for rapid fixation, decalcification, antigen
biology, vol 666. Springer, Heidelberg, pp retrieval, cryosectioning, and immunostaining.
327–343 Int J Cell Biol 2016:7076910
5. Kurth TB, De Bari C (2012) Immunostaining of
skeletal tissues. In: Helfrich M, Ralston H (eds)
Chapter 18
In Situ Hybridization in Mineralized Tissues: The Added

Value of LNA Probes for RNA Detection
G. Lignon, D. Hotton, A. Berdal, and A. Bolaños
Abstract
In situ hybridization (ISH) is one of the fundamental methods in developmental biology and neurobiology.
Their first ISH protocols were reported in 1969 (Gall and Pardue, Proc Natl AcadSci USA 63:378–83,
1969). Since several decades, ISH based on the specific hybridization of 100–2000 nucleotides long probes
enabled the localization of DNA/RNA sequences in tissues and cells with high cellular resolution. But
sometimes a limited sensitivity notably in mineralized tissues (Obernosterer et al., Nature Protocols
2:1508–14, 2007).
Here we describe a recent improvement of in situ hybridization efficiency by applying nucleotide locked
nucleic acid (LNA)-incorporated oligodeoxynucleotide probes (20 LNA/DNA nucleotide probes) essen-
tially used for noncoding miRNA and messenger RNAs.
Key words RNA probe, LNA/DNA probe, Specificity, Dental mineralized tissue
1 Introduction
Different techniques have been used to detect DNA/RNA such as

(RT-q)PCR, Northern blotting, hybridization to microarrays,
cloning, and sequencing, as well as single-cell microRNA
(miRNA) detection with ISH highlighting the difficulty to detect
them [1]. However, the only one providing insights into cell level
and localization is in situ hybridization [2].
Several noncoding RNAs notably for microRNA are expressed
at very low level and have a very short length. Visualization of these
noncoding RNA species in cells and tissues is an important tool to
help understand their roles at the cellular level.
In parallel with the discovery and characterization of novel
RNA species, novel tools to detect the RNA species in situ have
been invented and developed. Locked nucleic acid (LNA)-based
probes are the general probe technology used for miRNA detection
G. Lignon and D. Hotton contributed equally to this work.
181
182 G. Lignon et al.
Fig. 1 Localization of amelogenin mRNAs by in situ hybridization with digoxigenin RNA and LNA/DNA probes
with the same protocol on paraffin sections of mandibles of wild-type mice 11 days old. a ameloblasts,
o odontoblasts and *osteoblasts. RNA probes: in vitro transcription reaction: AMELX sense and antisense RNA
probes were prepared from rat amelogenin full-length 1, 1 kb cDNA (NM_007742.3) subcloned into Bluescript
plasmid and labeled with digoxigenin-dUTP by in vitro transcription using T7 and/or T3 RNA polymerase on
MAXIscript Kit (Ambion, USA). LNA probes: the final LNA/DNA digoxigenin probe for the same rat amelogenin
mRNA sequence was designed incorporating locked nucleic acid (LNA)-modified bases (sequence—gaggtgg-
taggggcatagcaaaa—) by Exiqon, Vedbaek, Denmark
in tissues and cells which enabled a significant profit in their detec-

tion (Fig. 1). Additionally, these probes have also been used in
cancer therapy in antisense strategy [3]. These probes are the gold
standard for miRNA detection [4].
LNA oligonucleotides are bicyclic RNA analogs which are
conformationally locked in a C3-endo N-type sugar conformation
by a 2’O and 4’C methyl bridge [5–7]. They exercise an unprece-
dented high affinity for their complementary DNA or RNA targets
molecules with good stability and aqueous solubility [6], allowing
for increased hybridization temperatures and highly stringent
rinsings.
Here we present an optimized protocol to detect and localize
mRNA expression of several genes. We successfully applied this
method in the paraffin dental tissue sections. Our method is char-
acterized by high sensitivity, specific histological detection of
mRNAs, and excellent morphology in mice maxillary tissue by in
situ hybridization with RNA probes and LNA/DNA probes
(Fig. 2).
LNA Probe for RNA Detection 183
LNA probes Riboprobes
Base Base Base

O O O
O O O
O O O O OH
O P O– O P O– O P O–
LNA DNA RNA
Fig. 2 Schematic diagram representing the structure of the LNA probes (left) and
of the riboprobes (DNA and RNA, right)
2 Materials
2.1 Deparaffinization 1. Clearene (Leica, France) (see Note 1).

and Rehydration 2. Ethanol.
3. Phosphate buffered saline: A liter stock of 10 PBS can be
prepared by dissolving 80 g NaCl, 2 g KCl, 14.4 g Na2H-
PO4·2H2O, 2.4 g KH2PO4, and 800 mL of Milli-Q H2O. After
complete mixing, top up final solution to 1 L. On dilution, the
resultant 1 PBS should have a final concentration of 137 mM
NaCl, 10 mM phosphate, 2.7 mM KCl, and a pH of 7.4 (see
Note 2).
Dispense the solution into aliquots and sterilize by autoclaving
(20 min, 121 C, liquid cycle). Store at room temperature.
2.2 Hybridization Standard precautions are made to ensure that all solutions and
Solutions equipment used are nuclease-free. Prepare all the solutions with
RNase-free water (Milli-Q H2O Synthesis Purification System or
similar) (see Note 3). Make sure that the work areas are RNase-free.
Always use disposable gloves (see Note 4).
1. Proteinase K treatment. Proteinase K buffer: Tris–HCl 50 mM
with 10 mM CaCl2 pH 8. Dissolve 121.14 g Tris in 800 mL
Milli-Q H2O. Adjust pH to 8.0 with the appropriate volume of
1MHCl. The final volume is 1 L Milli-Q H2O. Dissolve
110.98 g CaCl2 in 800 mL Milli-Q H2O. Add Milli-Q H2O
to a final volume to 1 L, and mix 50 mL Tris–HCl 1 M and
10 mL CaCl21M. Bring final volume to 1 L Milli-Q H2O and
adjust pH to 8.0. Autoclave and store at room temperature.
Proteinase K concentration stock solution: 20 mg/mL in Milli-

Q H2O at 20 C. Proteinase K concentration solution:
50 μg/mL.
2. Glycine buffer 2%: Dissolve 200 mg glycine in 100 mL 1
PBS. Autoclave and store at room temperature.
3. Postfixation: Make freshly 4% paraformaldehyde in 1 PBS
pH 7.4. For 800 mL of 4% formaldehyde (Electron Micros-
copy Science, USA), add 100 mL of 32% formaldehyde to a
glass beaker on a stir plate in a ventilated hood. Add 700 mL of
1 PBS pH 7.4. Verify the pH to approximately 7.4. The
solution can be aliquoted and frozen (20 C).
4. Triethanolamine (TEA) treatment. Triethanolamine 2 M: Dis-
solve 29.8 g triethanolamine in 100 mL Milli-Q H2O, pH 8.0
with HCl. Make 250 mL of 0.1 M triethanolamine solution
pH 8.0 by mixing 12.5 mL of 2 M triethanolamine into
237.5 mL Milli-Q H2O in a glass dish.
5. Acetic anhydride 99.8–100.5% AR (Merck, VWR, France).
6. Hybridization buffer: For 10 mL, mix in a Falcon tube 0.6 mL
5 M NaCl stock, 0.2 mL 1 M Tris–HCl stock pH 8.0, 0.1 mL
1 M NaPO4 stock pH 8.0, 0.1 mL 0.5 M EDTA stock, 5 mL
deionized formamide (Bio Basic Inc., #FB0211), 1 g dextran
sulfate (Euromedex, France), 0.2 mL 50 Denhardt stock
(Euromedex, France), 0.2 mL 0.5 mg/mL tRNA stock, and
3.7 mL H2O. Hybridization buffer without formamide can be
stored at 20 C (see Note 9).
2.3 Probes 1. RNA probes: Sense and antisense RNA probes were prepared
from full-length cDNA subcloned into Bluescript plasmid and
labeled with digoxigenin-dUTP by in vitro transcription using
T7 and Sp6 or T3 RNA polymerase on MAXIscript Kit
(Ambion, USA).
2. LNA probes: The final LNA/DNA digoxigenin probe for the
same mRNA interesting sequence was designed incorporating
locked nucleic acid (LNA)-modified bases by Exiqon, Vedbaek,
Denmark. The lyophilized LNA/DNA was suspended to
50 μg/mL with nuclease-free water and stored in 10–20 μL
aliquots at 80 C. Two negative controls were systematically
used: scrambled LNA/DNA (scramble control probe). No hits
of >70% homology to any sequence in any organism in the
NCBI database and omission of the LNA/DNA or RNA
digoxigenin probes.
2.4 Post- 1. Stringency washing (see Note 11): Make 20 SSC liter stock
hybridization Solutions solution: mix 175.32 g of sodium chloride (3 M) and 86.02 g
of trisodium citrate in 1 Milli-Q H2O liter (300 mM), and
adjust to pH 7.0 with 1 M HCl.
2. NTE buffer (140 mM NaCl, 50 mM Tris pH 7.5, 5 mM

EDTA): 2.8 mL 5 M NaCl, 5 mL 1 M Tris–HCl pH 7.5,
1 mL 0.5 M EDTA pH 8.0.
2.5 Hybridization 1. Blocking buffer: For 100 mL blocking buffer, dissolve 2 g

detection blocking reagent (Roche, Mannheim, Germany) in 1 PBS,
and add 20% heat-inactivated sheep nonimmune serum (heat
to 70 C for 30 min) (CliniSciences, France) and 0.1% Tween
(Euromedex, France).
2. Antibody conjugate: Dilute the AP-conjugated sheep anti-DIG
Fab fragment (Roche, Mannheim, Germany) in blocking buffer.
3. Washing solution: Add 100 μL Tween 20 in 1 PBS freshly.
4. Enzyme-substrate reaction: NBT/BCIP tablets (Roche, France).
5. Levamisole stock: Prepare 100 mM stock solution by adding
10 mL Milli-Q H2O to 250 mg levamisole. Add 200 μL of the
stock solution to 10 mL 0.1 M Tris–HCl, pH 9.5 (20 C) with
0.1 M NaCl.
2.6 Special 1. Water bath or slide incubation chamber.

Equipment 2. Dry bath incubator for 80–120 C.
3. Microcentrifuge.
4. Observations, image processing, and photography: a Leica
DMRB microscope (Leica Microscopy Systems, France) with
a camera. For image processing, ImageJ is used.
2.7 Other Essential 1. Staining dish with lid and slide rack. #21078 Ted Pella Inc.
Materials 2. Superfrost® Plus Microscope Slides (Thermo Scientific,
France).
3. Cover slips special 24 50 mm (Thermo Scientific, France).
4. DPX mounting.
3 Methods
3.1 Tissue Specimen The slides containing 5 μm-thick fixed tissue sections of decalcified
samples mice embedded in paraffin using standard procedure (see
Note 5).
3.2 Pretreatments Carry out all procedures at room temperature unless otherwise
Before Hybridization indicated.
1. Remove the wax from the paraffin tissue sections with Clearene
(Leica, France) 5 min twice.
2. Rehydrate the sections with ethanol 100% twice 3 min, ethanol
80% 3 min, and ethanol 70% 3 min.
3. Wash in phosphate buffered saline (PBS) 1 in Milli-Q H2O.

4. Proteinase K treatment: final concentration: 50 μg/mL in
200 mL Tris–HCl 50 mM, 10 mM CaCl2, pH 8 at 37 C,
5 min (see Note 6).
5. The reaction was stopped by wash with glycine buffer 2%.
6. Postfixation in freshly made 4% paraformaldehyde (Sigma-
Aldrich) solubilized (w/v) in 0.1 M phosphate buffer
(pH 7.4) for 20 min at room temperature (RT) (see Note 7).
7. Rinse PBS 1 in Milli-Q H2O, 2 min.
8. Kept on a stirrer in the fume hood, freshly prepare 400 mL of
0.1 M triethanolamine–HCl buffer (pH 8.0).
9. Primary incubation of the slides 10 min in 200 mL of 0.1 M
triethanolamine–HCl buffer (pH 8.0).
10. Discard the bath in a proper liquid can.
11. For the second incubation, add 625 μL of acetic anhydride into
the TEA buffer, and briefly stir. Reduce the speed of the stirrer,
and put the slides into the TEA (we use a clamping system with
a support stand to keep the slide rack elevated above the stir
bar) (see Note 8).
12. Continue to stir slowly for 10 min.
13. Rinse the slides twice in 1 PBS Milli-Q H2O for 2 min each.
3.3 Hybridization: Prepare RNA probe with in vitro transcription reaction (Ambion,
Day 1 USA) on ice as follows:
3.3.1 RNA Probe – 500 μL RNAse-free microfuge tube
Preparation and Storage – 4 μL 5 transcription buffer
– 1 μg DNA template
– 1 μL RNase out
– 2 μL digoxigenin riboprobe labeling mix (Roche, Germany)
– 2 μL RNA polymerase T7, RNA polymerase T3, or SP6 RNA
polymerase (20 U/μL)
– Milli-Q H2O to 20 μL
3.3.2 Procedure Incubate reaction at 37 C for 2 h with agitation.

Destroy DNA template by adding 2 μL RNAse-free DNAse I
(1 U/μL), and incubate reaction at 37 C for 15 min.
Centrifuge at >13,000 g for 30 min at 4 C to pellet.
Rinse pellet gently with 500 μL ice-cold 70 % ethanol.
Invert once.
Re-pellet by centrifugation at >13,000 g for 10 min.
Decant and air-dry carefully (~5 min).

Do not over dry.
Resuspend probe with 20 μL Milli-Q H2O and quantify.
Then dilute all probes in hybridization buffer in PCR tubes.
Prepare hybridization mixture with LNA/DNA probe by adding
1–2 μL of labeled LNA/DNA probe per 200 μL of hybridiza-
tion buffer (~30 ng/μL ¼ original concentration probe) and
with RNA probe by adding 6 ng/μL RNA probe/hybridiza-
tion buffer.
Heat at 95 C for 2 min in a PCR block to denature the RNA probe
and LNA/DNA probe. Chill on ice immediately to prevent.
Put 20–50 μL diluted probes per section.
Lay the slides down in a humidity chamber overnight (16–20 h) at
55 C for all probes (see Note 10).
Ensure that the slides do not touch each other, as this may lead to
mixing of the hybridization solutions.
3.4 Day 2: Post- In the post-hybridization treatments, Milli-Q water is not neces-
hybridization Rinsing sary; use distilled water only.
1. Wash 5 SSC RT by 15 min (see Note 11).
2. Wash (5 SSC, 50% formamide and 0.1% Tween) twice for
15 min at 55 C.
3. Wash (2 SSC+ 0.1% Tween) at RT by 15 min.
4. Wash (1 SSC + 0.1% Tween) at RT by 15 min.
5. Wash (0.1 SSC + 0.1% Tween) at RT by 15 min.
6. Rinse the slides in 1 PBS at room temperature for 30 min
twice.
7. Apply the blocking buffer (2% Boehringer blocking reagent
+20% nonimmune serum + distilled water) directly onto the
slides, and incubate the slides on a humidified platform over-
night at 4 C.
8. RNAse (5 mg/mL stock) step for riboprobes (not for LNA/
DNA probes). Rinse NTE (0.3 M NaCL, 10 mM Tris–HCl
pH 8, 5 mM EDTA for 50 μg/mL RNAse). NTE+ RNAse
(1 μL in 250 mL at 20 μg/mL) 10 min at 37 C. Rinse NTE
and 1 PBS.
3.5 Immunocyto- 1. Without rinsing discard the washing buffer, and apply
chemical Revelation 30–50 μL of anti-DIG solution directly onto the slides. Dilute
1/800 the anti-DIG antibody in the blocking solution (see
Note 12).
2. Make sure that the solution uniformly covers all sections, and
incubate the slides for 1 h at room temperature or 4 C over-
night at room temperature in humidity chamber (see Note 13).
3. Rinse twice in 1 PBS by 5mn at room temperature.
3.6 Detection Prepare the solution shortly before use.

Add 1 tablet to 10 mL distilled water to prepare a ready-to-use
staining solution 0.4 mg/mL NBT, 0.19 mg/mL BCIP in
100 mM Tris buffer 50 mM MgSO4, pH 9.5 with 2 mM levami-
sole. Check pH and filter the final solution.
Incubate together on slide for 1–4 days in the dark at RT, and
check the signal development under a microscope.
After color development and counter coloration (hematoxy-
lin), wash the slides in distilled water, and dehydrate by passing
through a series of alcohols (50, 75, 95, and 100%).
Mount using DPX and analyze.
4 Notes
1. Clearene solvent: The Clearene is a selected blend of terpenes

formulated specifically to replace xylene and toluene in histol-
ogy and cytology laboratory procedures. It is less oily than
other solvent substitutes and therefore dries quickly and does
not leave a residue. Clearene penetrates tissue rapidly without
overhardening and is safe to use with all open and closed system
tissue processors. It is miscible with alcohols, paraffin, and
mounting media. It contains an antioxidant to prevent stain
fading.
2. The pH of the 10 PBS stock is might be approximately 6.8
but when diluted to 1 PBS should change to 7.4. When
making buffer solutions, it is good practice to always measure
the pH directly using a pH meter. If necessary, pH can be
adjusted using hydrochloric acid or sodium hydroxide [8].
3. Milli-Q-treated water: We use water Milli-Q (Millipore Corpo-
ration, type 1 water) for all solutions before hybridization.
Water with an ultrafiltration/deionization cartridge (BioPak)
is suitable for genomic applications (quality at least equivalent
to DEPC-treated water, RNAse, DNAse, and nuclease-free).
4. Maintain an RNase-free environment during all steps in the in
situ hybridization procedure: Use autoclaved or sterile buffers
and heat-treated glassware, including cover glasses, and wear
gloves to minimize RNase contamination.
5. Complete decalcification of murine mandibles in 4.13% ethy-
lenediaminetetraacetic acid (EDTA) generally requires from
1 to 6 weeks for 11- to 120-day-old mice, respectively. To
accelerate the decalcification for in situ hybridization histologi-

cal analysis, we are now routinely using a microwave oven.
Microwave irradiation offers a means of delivering energy
directly to the specimen while providing control over the
amount of heat generated (30 C). Several studies [9] have
confirmed the efficiency of this method (good preservation of
the bone and dental tissues after microwave decalcification).
6. The proteinase K (PK) treatment step is critical for mineralized
tissues. Proteinase K is used for the proteolytic digestion of
paraffin-embedded, formalin-fixed mineralized tissues for trea-
ted crossed reactions with all cellular components. The PK
treatment increases the permeability and thereby the hybridiza-
tion efficiency. Insufficient digestion will reduce hybridization
signal, and over-digestion will result in poor tissue morphol-
ogy, making localization of the hybridization signal very diffi-
cult. The extent of proteinase K treatment is determined with
the proteinase K titration experiment concentration [10].
7. Short postfixation formaldehyde is required for the mineralized
tissue sections to remaintain the sections after PK digestion.
8. The TEA treatment acetylates the positively charged amino
groups within the tissue that may lead to a nonspecific binding
of the probe. The acetic anhydride in water produces a radical
acetyl which is a nucleophilic compound acting with the active
site of nucleases present inducing their inhibitions. Even with-
out TEA treatment, the level of background staining of the
mineralized tissue was present. In addition, acetic anhydride is
toxic; this solution should therefore be prepared in a fume
hood and disposed of properly.
9. The hybridization buffer is the same for LNA/DNA probe and
RNA probe [11]. Hybridization buffer can be prepared in large
batches, frozen in aliquots without formamide at 20 C.
10. The hybridization temperature is a critical parameter. Its
requires optimization, depending on the percentage of bases
in the target sequence of the probe used. An important factor is
the amount of cytosine and guanine in the sequence. In LNA
detection Exiqon supplies the Tm of every probe they distrib-
ute; generally the incubation temperature is 30 C below the
Tm of the probe.
We have obtained an optimal incubation temperature (55 C),
the same for all probes in our protocol.
11. The saline sodium citrate (SSC) buffer is used as a post-
hybridization buffer to control stringency for washing steps
in this protocol. Stringency is the solution capacity to act on
LNA/DNA or RNA pairing for the double-strand formation
and their stability. This step aims to denaturate the nonspecific
pairing and thereby to reduce background.
12. Sheep anti-digoxigenin alkaline phosphatase conjugate (Roche

Diagnostics). Centrifuge briefly to eliminate aggregate. Do not
freeze this antibody because it will lose much of its activity.
13. The duration of revelation depends on the signal intensity
(initial amount of RNA few hours to 4 days). Stop the color
reactions by rinsing the slides with distilled water. Rinse in
bi-distillated water.
References
1. Gu X, Li A, Liu S, Lin L, Xu S, Zhang P, Li S, analogues of LNA (locked nucleic acids):

Li X, Tian B, Zhu X, Wang X (2016) Micro- Phosphorothioate-LNA and 20 -thio-LNA.
RNA124 regulated neurite elongation by tar- Bioorg Med Chem Lett 8(16):2219–2222
geting OSBP. Mol Neurobiol 53 7. Obika S, Morio K, Hari Y, Imanishi T (1999)
(9):6388–6396 Preparation and properties of 20 , 50 -linked oli-
2. Levsky JM, Singer RH (2003) Fluorescence in gonucleotide analogues containing 3’-O,4’-C-
situ hybridization: past, present and future. J methyleneribonucleosides. Bioorg Med Chem
Cell Sci 116(Pt 14):2833–2838 Lett 9(4):515–518
3. Stenvang J, Lindow M, Kauppinen S (2008) 8. Fritsch EF, Maniatis T (1989) Molecular clon-
Targetings of microRNAs for therapeutics. ing. In: Cold Spring Harbor Laboratory Press
Biochem Soc Trans 36(Pt 6):1197–1200 (ed) A laboratory manual, vol 3, 2nd edn. Cold
4. Urbanek MO, Nawrocka AU, Krzyzosiak WJ Spring Harbor, New York, p 12
(2015) Small RNA detection by in situ hybri- 9. Cunningham CD 3rd, Schulte BA, Bianchi
dization methods. Int J Mol Sci 16 LM, Weber PC, Schmiedt BN (2001) Micro-
(6):13259–13286 wave decalcification of human temporal bones.
5. Koshkin AA, Singh SK, Nielsen P, Rajwanshi Laryngoscope 111(2):278–282
VK, Kumar R, Meldgaard M, Olsen CE, Wen- 10. Jørgensen S, Baker A, Møller S, Nielsen BS
gel J (1998) LNA (locked nucleic acids): syn- (2010) Robust one-day in situ hybridization
thesis of the adenine, cytosine, guanine, protocol for detection of microRNAs in paraf-
5-methylcytosine, thymine and uracil bicyclo- fin samples using LNA probes. Methods 52
nucleoside monomers, oligomerisation, and (4):375–381
unprecedented nucleic acid recognition. Tetra- 11. Wilkinson DG, Nieto MA (1993) Detection of
hedron 54:3607–3630 messenger RNA by in situ hybridization to
6. Kumar R, Singh SK, Koshkin AA, Rajwanshi tissue sections and whole mounts. Methods
VK, Meldgaard M, Wengel J (1998) The first Enzymol 225:361–373
Chapter 19
Immunofluorescence Procedure for Developing

Enamel Tissues
Xu Yang and Elia Beniash
Abstract
Immunofluorescence (IF) labeling is a powerful technique that can provide a wealth of information on
structural organization, supramolecular composition, and functional properties of cells and tissues. At the
same time, nonspecific staining and false positives can seriously compromise IF studies and lead to
confusing or even misleading results. It is particularly true for the extracellular matrix component of
forming enamel. Here, we present an optimized IF protocol for developing enamel. Autofluorescence
blocking by Sudan Black B (SBB) and establishing of proper isotype controls lead to a significant artifact
reduction and improve reliability of the IF data.
Key words Autofluorescence, Enamel, Sudan Black B, Immunofluorescence, False positive
1 Introduction
Immunofluorescence (IF) is a well-established technique which like

other immunostaining methods utilizes specific affinity between
antibodies and antigens to identify macromolecules of interest. In
typical immunostaining procedures, antibodies raised against the
molecule of interest (primary antibody) are applied to the sample,
and the bound antibodies are then visualized by incubation with
antibodies against the primary antibody, which are conjugated with
a probe (secondary antibody). In the case of IF, this probe is a
fluorophore which can be excited by irradiation of the sample with
light of a certain wavelength. Sometimes, primary antibodies con-
jugated with a fluorophore may also be used. Despite the simplicity
of the technique, there are numerous factors that affect the quality
and fidelity of the data. Common problems encountered by the
researchers are high background signal due to autofluorescence or
nonspecific binding of the antibodies [1, 2]. SBB has long been
used to eliminate autofluorescence in histology studies, although
the exact mechanisms are unknown. It is shown to dramatically
reduce background signals in tissue sections [3–11]. Another
191
192 Xu Yang and Elia Beniash
problem which can significantly affect IF analysis is nonspecific

binding of the antibodies which usually accounted for by using
isotype controls [1, 12]. Forming enamel is autofluorescent and
tends to interact nonspecifically with antibodies, which poses chal-
lenges for IF studies of this tissue [13]. Here, we present an
optimized IF protocol for developing enamel.
2 Materials
All solutions are prepared using deionized water. Most protein-

containing solutions, such as antibodies and sera, have to be ali-
quoted and stored at 20 C. Use within 3 days once thawed. If
not specially noted, the chemicals are stored and procedures con-
ducted at room temperature.
2.1 Fixation, 1. Fixation solution: 4% paraformaldehyde, 10 mM phosphate

Decalcification, buffer, and 150 mM NaCl, pH 7.2–7.4.
and Processing 2. Decalcification solution: 0.1 mM ethylenediaminetetraacetic
acid (EDTA), 10 mM phosphate buffer, and 150 mM NaCl,
pH 7.2–7.4.
3. Tissue processor and embedding station for histology.
4. Xylene, 100% ethanol, 95% ethanol, 70% ethanol.
2.2 Immunostaining 1. Tris–HCl buffer saline containing Tween (TBST): 50 mM

Tris–HCl, 150 mM NaCl, and 0.05% Tween 20, pH 7.2–7.4.
2. Antigen retrieval solution 1: 0.25% trypsin and 0.2% EDTA in
Hanks’ Balanced Salt Solution (common digestion solution for
cell culture) or 10 mg/mL proteinase K in TBS. Aliquot by
1 mL and store at 20 C.
3. Antigen retrieval solution 2: 10 mM sodium citrate buffer,
pH 6.0. Add 2.94 g Tri-sodium citrate (dehydrate) into 1 L
water and then adjust acidity.
4. Blocking solution: 10% serum of secondary antibody host ani-
mal (e.g., donkey serum) mixed with 2.25% IgG-free BSA,
0.1% Triton X-100, 0.14% glycine, 0.23% casein, 0.1% gelatin,
and 0.05% sodium azide in TBST at the proportion. Aliquot by
1 mL and store at 20 C (see Note 1).
5. Antibody dilution solution: ½ blocking solution.
6. Antibody solution: Self-made or purchased primary antibody is
aliquoted and stored at 80 C. Pre-bleed sera or normal IgGs
are used as isotype controls (see Note 2). Secondary antibody
with particular fluorophore is purchased.
Immunofluorescence Procedure for Murine Developing Enamel Tissues 193
7. SBB solution: 1.5% SBB in 70% ethanol. Add 0.3 g SBB into
20 mL 70% ethanol, mix thoroughly, and store at 4 C (see
Note 3).
8. Counterstain solution: DAPI 0.5 μg/mL in TBST (see
Note 4).
3 Methods
3.1 Fixation, 1. Mandibles of mice or rats are promptly dissected after euthana-
Decalcification, sia and immediately immersed in fresh 4% paraformaldehyde in
Embedding, and PBS (at least x20 of the sample volume) and kept at 4 C for
Sectioning of the 24 h.
Samples 2. Fixative solution is removed and replaced with 0.1 M EDTA in
PBS (ten times the sample volume) at 4 C, and the decalcifi-
cation medium is replaced every other day. Decalcification
procedure is complete when the sample becomes soft and
easy to bend. For normal 4-week-old mice and 4-week-old
rats, decalcification will take roughly 1 and 2 weeks,
respectively.
3. After decalcification the samples are washed with PBS (20 times
the sample volume) 30 min for three times.
4. The samples are dehydrated in 30, 50, and 70% ethanol gradi-
ent (20 times volume) for 30 min each step. The following
steps are conducted using paraffin processer. Set the cycle as
70% ethanol (37 C) 1 h, 95% ethanol (45 C) 1 h twice, 100%
ethanol (45 C) 1 h for three times, xylene (45 C) 1 h for three
times, and paraffin (65 C) 1 h for three times.
5. After dehydration and infiltration, the samples are embedded in
fresh wax with side facing a bottom of the mold. The buccal
surface should be cleaned from muscles and connective tissues
prior to the dehydration and embedding procedures.
6. The paraffin blocs are stored at 4 C and sectioned using a
routine histological microtome to the thickness of 6–12 μm.
3.2 Antigen Retrieval 1. The sections are deparaffinized as follows: xylene 4–5 min 3,
100% ethanol 2 min 2, 95% ethanol 2 min, 70% ethanol
2 min, and distilled water 2 min 2.
2. Two antigen retrieval methods (depending on the antibody
and the species, one can be better than the other):
(a) The sections are encircled with a water-proof pen. After
the water-proof dye dries, the sections are washed by
placing droplets of TBST over the sections for 3 min 3
times. Preheat antigen retrieval solution 1 and the slides in
a humidified chamber for 10 min at 37 C. Place antigen
retrieval solution 1 over the sections, and incubate for

5–20 min at 37 C (optimal incubation time has to be
determined for each individual antibody). Wash the sec-
tions with TBST 1 min 3 times.
(b) Preheat the antigen retrieval solution 2 in a water bath to
96–100 C. Transfer deparaffinized slides into the jar,
close the lid, and incubate for 5–20 min at the same
temperature (optimal time to be decided for each individ-
ual antibody). Let the jar cool down, take out the slides,
and draw water-proof circles around the sections. Wash
with TBST 3 min 2 times.
3.3 Antibody 1. Make full blocking solution and add around 50 μL onto each
Incubation section. Incubate for 1 h at room temperature.
2. The blocking solution is removed with a pipette or is shaken off
the slide. The primary antibody solution (around 50 μL) is
placed over the section and incubated in a humidified chamber
overnight at 4 C.
3. Wash with TBST for 4 min 6 times.
4. Incubate with diluted secondary antibody for 45–60 min at
room temperature.
6. If double or triple labeling is required, incubate with blocking
solution for 30 min, and start from step 2 again.
7. Incubate with SBB solution for 20–30 min.
8. Wash away SBB with large volume of TBST and another TBST
wash for 4 min.
9. Counterstain with DAPI solution for 5 min (see Note 4).
11. Mount the slide with commercial fluorophore protective
mounting medium or just observe in PBS without mounting.
4 Notes
1. Casein is hard to dissolve in water. Weigh 0.6 g casein powder,

add 100 mL water, and stir at 60 C for 5 min. Add 1 M NaOH
solution drop by drop until all casein powder dissolves, and stir
for another 5 min. Let the solution cool down, add 12 mL 10
PBS, make total volume 120 mL, and adjust pH to 7.2–7.4
(the casein solution might still appear cloudy). Place it at 4 C
overnight to let the particulate phase to sediment at the bot-
tom. Collect only 100 mL of the supernatant, and mix it with
100 mL PBS solution containing 5% IgG-free BSA, 0.2%
Immunofluorescence Procedure for Murine Developing Enamel Tissues 195
Triton X-100, 0.3% glycine, 0.2% gelatin, and 0.06% sodium

azide. Aliquot the mixture at 1 mL per tube and store at
20 C. To make the full blocking solution, take one tube of
the mixture out, add 111 μL of serum from the host animal of
the secondary antibody, and thoroughly mix it.
2. Our observations indicate that normal sera have a much higher
affinity to enamel matrix than to dentin and bone. Importantly,
sera from different species differ greatly in their affinity toward
the enamel matrix. For example, sera from rodent hosts (guinea
pig and rat) have a much lower affinity to enamel matrix of mice
than non-rodent hosts, such as sheep, goat, and rabbit [13].
Based on the analysis of interactions between different sera and
the enamel matrix, it is highly recommended that the normal
sera isotype control is used at the dilutions <1/100 for rat and
guinea pig antibodies and < 1/500 for other sera, to avoid false
positives.
Affinity-purified IgGs from most species have very low
affinities toward the enamel matrix at concentrations below
5 μg/mL. However, polyclonal rabbit IgG shows a fairly strong
affinity toward the enamel matrix even when its concentration
is as low as 1 μg/mL [13]. At the same time, monoclonal rabbit
IgGs (5 μg/mL) targeted at intracellular proteins do not show
any binding to enamel matrix. Thus, isotype control is highly
recommended when using rabbit polyclonal IgG at concentra-
tion close to or higher than 1 μg/mL.
3. A 1.5% SBB will be highly supersaturated in 70% ethanol. Add
0.3 g SBB into 20 mL 70% ethanol, mix thoroughly for at least
10 min, and then let it sit for 10 min to let the undissolved
material to precipitate. Use a 20 mL syringe to collect the
supernatant, attach a 0.2 μm or 0.45 μm filter to the syringe,
wrap the assembly with aluminum foil, and cap the filter tip
with parafilm. Keep the syringe at 4 C and it should be stable
for months. Each time before use, take it out 5–10 min earlier
to warm up, then pump out the solution through filter, and
drop directly onto the sections. The color of SBB solution
might turn slightly pink or red after incubation, which is very
common.
4. DAPI staining will be compromised a little by SBB. You can try
higher concentrations or longer incubation times if stronger
signals are required. Normally, the concentration and time
listed in this protocol are good for most situations.
References
1. True LD (2008) Quality control in molecular 2. Burry RW (2011) Controls for immunocyto-
immunohistochemistry. Histochem Cell Biol chemistry. Journal of Histochemistry & Cyto-
130:473–480. https://doi.org/10.1007/ chemistry 59:6–12. https://doi.org/10.
s00418-008-0481-0 1369/jhc.2010.956920
3. Erben T, Ossig R, Naim HY, Schnekenburger J 8. Romijn HJ et al (1999) Double immunolabel-

(2016) What to do with high autofluorescence ing of neuropeptides in the human hypothala-
background in pancreatic tissues - an efficient mus as analyzed by confocal laser scanning
Sudan black B quenching method for specific fluorescence microscopy. J Histochem Cyto-
immunofluorescence labelling. Histopathol- chem 47:229–236
ogy 69:406–422. https://doi.org/10.1111/ 9. Sun Y et al (2011) Sudan black B reduces auto-
his.12935 fluorescence in murine renal tissue. Arch
4. Kajimura J, Ito R, Manley NR, Hale LP (2016) Pathol Lab Med 135:1335–1342. https://
Optimization of single- and dual-color immu- doi.org/10.5858/arpa.2010-0549-OA
nofluorescence protocols for formalin-fixed, 10. Viegas MS, Martins TC, Seco F, do Carmo A
paraffin-embedded archival tissues. J Histo- (2007) An improved and cost-effective meth-
chem Cytochem 64:112–124. https://doi. odology for the reduction of autofluorescence
org/10.1369/0022155415610792 in direct immunofluorescence studies on
5. Nakata T et al (2011) Simultaneous detection formalin-fixed paraffin-embedded tissues. Eur
of T lymphocyte-related antigens (CD4/CD8, J Histochem 51:59–66
CD57, TCRbeta) with nuclei by fluorescence- 11. Yang Y, Honaramooz A (2012) Characteriza-
based immunohistochemistry in paraffin- tion and quenching of autofluorescence in pig-
embedded human lymph node, liver cancer let testis tissue and cells. Anatomy research
and stomach cancer. Acta Cytol 55:357–363. international 2012:820120. https://doi.org/
https://doi.org/10.1159/000329487 10.1155/2012/820120
6. Neo PY, Tan DJ, Shi P, Toh SL, Goh JC 12. Tan NCW, Tran H, Roscioli E, Wormald PJ,
(2015) Enhancing analysis of cells and proteins Vreugde S (2012) Prevention of false positive
by fluorescence imaging on silk-based bioma- binding during immunofluorescence of Staph-
terials: modulating the autofluorescence of silk. ylococcus aureus infected tissue biopsies. J
Tissue Eng Part C Methods 21:218–228. Immunol Methods 384:111–117. https://
https://doi.org/10.1089/ten.TEC.2014. doi.org/10.1016/j.jim.2012.07.015
0209 13. Yang X, Vidunas AJ, Beniash E (2017) Opti-
7. Oliveira VC et al (2010) Sudan Black B treat- mizing immunostaining of enamel matrix:
ment reduces autofluorescence and improves application of sudan black b and minimization
resolution of in situ hybridization specific fluo- of false positives from normal sera and IgGs.
rescent signals of brain sections. Histol Histo- Front Physiol 8:239. https://doi.org/10.
pathol 25:1017–1024 3389/fphys.2017.00239
Chapter 20
Silver-Albumin Tissue Staining Protocol to Visualize

Odontogenesis in Whole Embryos
Julia C. Boughner and David M. L. Cooper
Abstract
Visualizing tooth organs from their earliest inception as they actually appear in three dimensions has, until
recently, been difficult due to the technical obstacle of imaging these tiny, translucent, low-density
embryonic craniodental tissues. Related to this obstacle, quantifying craniodental morphology has been
confounded by the time consuming need to physically section and then digitally photograph and recon-
struct these images of tissues into 3D volumes. Here we provide a simple solution in the form of an
overnight silver albumin tissue stain for whole embryos. Because it is differentially absorbed by embryonic
tissues, this stain generates the contrast needed to detect and visualize unmineralized dental tissues. Stained
specimens can be scanned using either desktop or synchrotron micro-computed tomography systems,
generating digital 3D datasets of whole embryos that can immediately be used to assess dental morphology
and histology. Craniodental structures can then be measured with high precision and accuracy using 3D
image analysis software.
Key words Contrast agent, Tooth development, Micro-computed tomography, 3D imaging, Mor-
phogenesis, Virtual histology, Synchrotron
1 Introduction
Most of what we know about the classic stages of odontogenesis

and the structure of the tooth organ and the dental lamina has been
discovered via traditional histology techniques using tissue sections
(e.g., [1–3]). Typically, research reports and textbooks show the
developing tooth organ in coronal view because this 2D plane best
illustrates the organ’s various layers and structures. However, this
classic coronal view is problematic because the tooth organ appears
to be spherical when in fact it is tubular, as clearly shown by more
recent, meticulous 3D reconstructions of 2D histological sections
through the dental lamina [4, 5].
The Herculean task of reconstructing odontogenesis in 3D
using serial histology was undertaken because of the paucity of
methods to directly image in 3D the developmental morphology
197
198 Julia C. Boughner and David M. L. Cooper
of tooth organs. These reconstructions are very useful yet very time
and energy intensive. Embryos must be sliced, the tissue sections
slide-mounted and digitally photographed, and the 2D images
digitally assembled into a virtual 3D volume. As such, the recon-
structions may lack precision because of error introduced during
the multistep process. Also, the labor-intensive nature of this recon-
struction process limits which structures are included: thus dental
tissues are studied in isolation, out of context of surrounding jaw
tissues. Triaging non-dental structures precludes visualizing tooth
and jaw structures forming relative to each other. Yet these more
comprehensive image data are of particular interest and necessity to
those studying the developmental integration of dental and facial
phenotypes (e.g., [6–10]).
As a less destructive and more efficient alternative, we have
optimized a method to directly visualize embryonic dental and
adjacent facial tissue morphology in 3D. This overnight tissue
staining protocol penetrates intact whole embryos and selectively
impregnates a variety of tissue types including dental, skeletal,
nervous, and vascular structures. Micro-computed tomography
(μCT) scans of silver-stained embryos accurately capture develop-
mental morphology in fine detail as it actually appears in 3D. The
digital format of the scan data allows dental and neighboring facial
tissues to be measured with high precision and accuracy using the
3D measurement tools in most any 3D image analysis software
package, either proprietary (e.g., Amira, FEI) or open source
(e.g., Blender, Blender Foundation). This protocol contributes to
a growing toolkit of tissue contrast stains [11–13] that are helping
anatomists, morphologists, and developmental biologists to visua-
lize and quantify the processes via which genotype is translated to
phenotype.
2 Materials
The following tools, tubes, solutions, or other materials do not

need to be sterile since this method aims to capture morphological
and not molecular data. Distilled water (dH2O) can be used for all
stock solutions. All steps are performed at room temperature except
where noted. Gloves are required for the protection of the user, not
the samples. Please dispose of used fixative and ethanol solutions as
per your institution’s waste management protocols (e.g., some
places have an ethanol recycling program). Protargol (silver albu-
min) solution must also be disposed of as a toxic waste (i.e., do not
flush down the sink; instead bottle for removal and treatment).
2.1 Embryo 1. Phosphate buffered saline (PBS) solution diluted to 1 (e.g.,

Collection 100 mL of 10 stock PBS solution in 900 mL dH2O). 1 PBS
can also be made from powder or solid tablets dissolved in
A Silver Stain to Visualize Odontogenesis in 3D 199
Fig. 1 Dissection setup with forceps (left) and plastic transfer pipette scoops (center bottom), ice pack and
petri dishes (center), and 7+ mL specimen tubes (right) all laid out on the stage of the stereo microscope
dH2O. Store 1 PBS in a clean, easily pourable plastic or glass

container. It is useful to aliquot (i.e., measure and pour) smaller
volumes of 1 PBS into standard clear, graded 50 mL tubes.
2. Paraformaldehyde (PFA) fixative solution diluted with 1 PBS
to 4% concentration (PFA can also be bought in 4% liquid form
or made from stock powder). If desired, a fixative solution of
4% PFA + 5% glutaraldehyde has been shown to minimize
tissue shrinkage [14] and is therefore the better fixative for
studying and quantifying morphology. To make this fixative
solution, add 5 mL of 50% glutaraldehyde to 45 mL of 4% PFA
(made in 1 PBS). Store at 4 C (39.2 F) if you plan to use
this solution within a few days (see Note 1).
3. Dissection tools: (a) Two fine-point forceps with tips that meet
and are sufficiently sharp to cleanly dissect mouse embryos
from the uterus and remove even the thinnest placental tissues
(i.e., amnion) from the surface of each embryo so that the
fixative solution can fully penetrate the body. (b) A
non-sterile plastic transfer pipette, cut diagonally either
1–2 cm from its narrowest end or across the opposite bulb
end, to create a “spoon” with which to scoop up embryos
without having to grab, pierce, or otherwise damage the soft
tissues and destroy embryo morphology (Fig. 1).
4. Dissecting stereo microscope (e.g., Olympus SZX10). A
gooseneck lamp is ideal for illuminating embryos for precise
dissecting work.
5. Small (~65 mm diameter) plastic or glass petri dishes (Fig. 1).
The exact size is not critical, but if the dish is very large (e.g.,
100+ mm diameter), then it is harder to corral the embryos
under the light source and dissect them out. However, too
small of a dish makes it harder and slower to manipulate the
tissues and extract the embryos.
6. Styrofoam box (e.g., used to ship antibodies) or lab ice bucket,

filled almost to the top with ice, for cold-storing tubes of
fixative and embryos once these are collected. Also, a frozen
ice pack placed on the microscope stage on which to place your
petri dish while dissecting out embryos. This ice pack is not
essential, but it is helpful to maintain tissues if the dissection
process is taking a longer time to complete.
7. Standard flat-top snap-cap (not twistable, for ease of single-
handed use) clear plastic 2.0 mL microcentrifuge tubes and
clear 7 mL polystyrene bijou twist top tubes (see Note 2).
8. If embryos need to be genotyped as well as scanned, then prior
to embryo collection, prepare permanent fine-tip markers for
tube-labeling, and 1.5 mL or 0.5 mL clear plastic tubes in
which to place tail clips, which can be cut off using forceps.
2.2 Embryo 1. 100% (anhydrous) ethanol solution for final washes, and ~99%
Dehydration ethanol solution that will be diluted with either 1 PBS or
dH2O for all other washes (~99% ethanol is simply less expen-
sive than anhydrous stock solution).
2. Graded clear polypropylene tubes 50 mL or larger in which to
make and store each concentration of ethanol solution.
3. For the initial 25% ethanol solution wash, combine 25 mL of
~99% ethanol in 75 mL of 1 PBS (or, if working with a 50 mL
tube, pour half the volume, 12.5 mL, of ethanol and top up
with PBS). For the next 50% concentration, combine 50:50
(or 25 mL:25 mL) of ~99% ethanol:1 PBS. For the 70 and
90% ethanol washes, combine 70 mL or 90 mL of ~99%
ethanol with dH2O. Use anhydrous ethanol for the final
100% concentration washes.
4. Bench coat, under pads (“diapers”), or paper towels to absorb
any spilled solutions.
5. Disposable non-sterile plastic transfer pipettes—or, if embryos
are very young and tiny, 200 or 500 μL pipetman with appro-
priate plastic pipette tips to suck up the wash solution from
within the 1.5 mL tube without sucking up the embryo.
6. Permanent markers (if possible, ethanol/chemical resistant
pens) with which to mark the ethanol concentration on each
50 mL tube and the specimen number on each 2 mL or
7 mL tube.
7. Waste containers for storage and disposal of used fixative and
ethanol solutions.
8. Refrigerator or cold room for overnight or longer-term cold
storage.
Fig. 2 (a) Standard histology oven set to 37 C, at the foot of which is a glass
histology trough filled with 99 mL distilled water, a bottle of Protargol-S powder
(Polysciences Inc., now available from Sigma) from which 1 gram of powder is
measured, and two histology cassettes into which embryos are placed before
the cassettes are put into the silver stain solution (1 g Protargol dissolved in
99 mL H2O). (b) Close-up of trough, cassettes, and jar of Protargol
2.3 Embryo Staining 1. Histology oven and thermometer to check correct internal
temperature (Fig. 2).
2. Parafilm, pencil, and eraser.
3. Glass histology staining troughs with glass lids (Fig. 2).
4. Distilled water.
5. Plastic tissue processing embedding cassettes with lids (Fig. 2)
(e.g., Shandon, Thermo Fisher) and a slotted “pore style” to
increase fluid circulation within the cassette. Cassettes that have
internal compartments are optional and only useful for proces-
sing very small specimens (e.g., mouse aged embryonic day
(E) 10). Standard cassette height (5 mm) will fit most embryos;
larger specimens may only fit a larger height of cassette
(10 mm).
6. Protargol/silver albumin powder (Sigma, silver proteinate
(Protargol) ~8% Ag basis 05495).
7. Lab scale, sensitive to at least three decimal places (1000 mg),
disposable or reusable weigh boat, clean metal or plastic scoop,
fume hood, lab coat, and gloves.
2.4 Embryo 1. Desktop μCT scanner (e.g., SkyScan 1172, Bruker) and its
Scanning operating software on a desktop computer; or a μCT config-
ured beamline (e.g., our local Biomedical Imaging and Therapy
[BMIT] beamline) at a synchrotron (e.g., the Canadian Light

Source [CLS], Saskatoon) (see Note 3).
2. Clear polypropylene (see Note 4) 1.5 mL conical microcentri-
fuge tube, plastic petri dish, parafilm, forceps, plastic transfer
pipette cut into “spoon,” 100% ethanol solution, extra 50 mL
tubes to hold solutions transferred from specimen tube.
2.5 Scan File Size Sufficient internal data storage for scan files that are each 4–10 GB
and Storage in size; external 500+ GB drive(s) for back up and transport among
computers, unless the computers are networked to each other and
to, for example, a RAID storage system. If a computer lacks suffi-
cient memory to render and collect measurement data from scan
image data, then external drives 1 TB or larger are useful because
the software can read the raw scan data directly from the external
drive (although this is not the ideal because it is a bit slower and less
stable), and post-processing image and measurements data can be
stored on the same drive.
2.6 Image Rendering 1. Computer workstation that runs a Windows operating system,
and Analysis with as large and as fast a central processing unit (CPU) and a
graphics processing unit (GPU) as possible.
2. Once loaded into the 3D image analysis software of your
choice, then lengths, widths, heights, etc. within a 3D scan
volume (e.g., of a whole embryo) can be measured in x, y, and z
plains. The data can be exported as a file that can be read in
spreadsheet-type software (e.g., MS Excel).
3 Methods
To avoid damaging tissue and destroying morphology, whenever

possible avoid touching the specimen (i.e., gently tip it from a tube
into another tube or dish), and use softer tools (i.e., blunt plastic
pipette, tip of a tightly twisted piece of Kimwipe) to manipulate and
position the embryo for scanning.
3.1 Making Up PFA Thaw frozen fixative (or remove from fridge, and keep the solution
and then Adding Glute: chilled on ice). To thaw fixative, fill a container with warm (not hot)
Day 0 water (or use a warm water bath), and place the tube of frozen
fixative solution into the water to gently thaw (see Note 1).
3.2 Embryo 1. Fill a medium-to-large foam shipping container or rubber

Collection and bucket with ice. Chill the fixative solution in the ice (e.g.,
Fixation: Day 1 depending on specimen number and age, use a volume of
50 mL, possibly more). Count out the required number of
specimen tubes: a flat-bottomed 2 mL tube will work for very
small/young embryos up to about E12; for embryos larger/
older than about E12 (see Note 5), a 7 mL tube offers a better
tissue/fixative ratio. Fill each tube to about ¾ full with cold

fixative. Keep the remaining volume on ice for later use. In the
same ice bucket, chill the media in which the uteri (see next
step, Subheading 3.2, step 2) will be placed and the embryos
will be dissected out. The next step assumes that the user has
been granted access to laboratory animals that are bred to
generate embryos for collection, staining, and scanning. This
protocol was developed using mouse embryos; however
embryos of chick, fish, frog, and other model species would
also work.
2. After being sacrificed (see Note 6), the abdomens of pregnant
dams (female mice) are opened using small surgical scissors and
forceps to hold the skin away from the internal organs and
avoid cutting into the uterus accidentally. Videos of similar
protocols are available online (e.g., Shea, K., Geijsen,
N. Dissection of 6.5 dpc Mouse Embryos. J. Vis. Exp. (2),
e160, doi:10.3791/160 (2007); YouTube channel “Protoco-
lExchange”). Once the abdomen is opened, locate the uterus.
This organ is recognizable by its general “Y” shape, with left
and right halves, as well as its “string of pearls” appearance,
where each “pearl” is an embryo and its corresponding placen-
tal and amniotic tissues. Using forceps and scissors, carefully lift
the uterus away from the abdomen and snip both far right and
left sides, as well as the base of the uterus, to separate it from
the body.
3. Place an ice pack on the stage of the dissecting microscope.
Place each top/bottom half of a petri dish on the ice pack, and
fill each half-dish halfway to the top with cold media. Remove
the entire uterus and place it in one media-filled half petri dish.
Working under the microscope using forceps, gently and slowly
tease open the opaque uterus tissues around each “pearl” (i.e.,
embryo). If possible, keep the thin translucent chorionic mem-
brane (i.e., the fluid-filled “balloon” around each embryo)
intact. Gently pinch a section of this membrane between for-
ceps; lift and move the entire embryo into the second media-
filled half petri dish. This second dish is better for fine dissec-
tion work because most of the blood from the uterus is left
behind in the first dish and doesn’t obscure your work. Gently
dissect away the outer chorionic and inner amniotic mem-
branes. Any amniotic membrane left around the embryo’s
body will impede penetration of the fixative and, later, the silver
stain.
4. Once the embryo is fully dissected out, there are two choices.
(1) If the embryo needs to be genotyped, then using forceps
cut off a small piece of the tail (for example), put this tissue
sample into an empty 0.5 or 1.5 mL micro centrifuge tube, and
freeze the sample at 20 C as soon as possible. To avoid DNA
degradation, ideally within 2 weeks, thaw and extract DNA

from this and any other samples. Store the extracted DNA in
solution form at 4 C to avoid DNA shearing due to freeze-
thawing. (2) Next—or if the embryo does not need to be
genotyped—gently use the plastic scoop (made by cutting the
end of a plastic transfer pipette) to move the embryo from the
petri dish to its specimen tube half-filled with fixative. Top up
the volume of fixative so that even if the tube is lying on its side,
there is minimal air left in the container, and the embryo will be
fully immersed. Label each tube on the lid and/or side with the
necessary specimen identifying details (e.g., stage, specimen
number, and collection date). Leave each tube on ice until it
is time to move the tubes to 4 C for overnight fixation and
storage. When tubes are moved into a fridge or cold room,
store tubes in a sealable plastic bag to prevent fixative leaking
onto other items. We recommend storing the tubes on slowly
moving rockers to increase fixative penetration. We also recom-
mend storing tubes on their sides to allow maximum move-
ment of fixative over maximum surface area of each embryo.
3.3 Embryo 1. The next day, at least 16 h after embryos were put into fixative,
Dehydration and remove the specimens from cold storage. Working over bench
Processing: Day coat, paper towels, or similarly absorbent covering, prepare the
2 (See Note 7) graded solutions of ethanol (as per Subheading 2.2, item 3).
When done, open the first specimen tube. If possible, pour out
the majority of the fixative into an appropriate waste container.
We recommend using an old 50 mL tube in case the specimen
is poured out along with the solution. It is much simpler to
retrieve the specimen from a 50 mL tube than a container
500 mL or larger in volume. When the 50 mL tube is full, tip
its waste contents into your main, larger waste container. Use a
plastic transfer pipette, or a “p200” (200 μL) pipetteman fitted
with the corresponding tip, to suck up all remaining fixative
solution, and eject this into the waste container. Once almost
all the fixative is removed, replace it with 1 PBS, filling the
tube to the top. Do this for each specimen. Ideally, place all
tubes on a rocker for the duration of each wash. Otherwise,
hand roll or gently tilt the tubes every minute or 2. Depending
on the size/age of the specimen, each solution “wash” is done
for 10–15 min. Typically, we do shorter (10 min) washes for
the first 1 PBS and first 100% ethanol solutions because these
are typically the most contaminated with previous solutions
(particularly the 1 PBS wash which will contain large traces
of fixative solution).
2. Embryos can remain stored in 100% ethanol at 4 C for weeks
without damage to tissues; however, the longer that tissues stay
in 100% ethanol, the more brittle and more prone to damage
they will be. We prefer to collect specimens 1–4 weeks before

we require them for scanning. An alternative is to store speci-
mens in 70% ethanol and process the embryos into 100%
ethanol the morning that you plan to silver-stain them.
3.4 Embryo Staining: 1. Heat a histology oven to 37 C.

Day 3 or Later 2. We use a 1% (wt/vol) stain solution. First, fill a glass histology
staining trough with 99 mL dH2O. Place the trough in the
histology oven. Next, weigh out 1 g of Protargol (silver albu-
min powder). Gently sprinkle the powder onto the surface of
the water. Do not stir: agitating the powder somehow disrupts
the stain’s activity for reasons that remain unclear to us and
others (see [6] for more discussion). Gently place a layer of
parafilm such that it completely covers and hangs over the
sides of the trough’s walls. Put the trough’s lid securely in
place, gently pressing the lid into the parafilm to make a rela-
tively airtight seal. Close the door of the histology oven to
maintain temperature, and allow the powder to fully dissolve
into solution (30 min or longer).
3. While waiting for the powder to dissolve, label with pencil (pen
may rub or wash off) a plastic tissue processing cassette with
each specimen number or other identifier. Place each embryo in
its own cassette, transferring as little of the ethanol solution as
possible.
4. Once the powder is dissolved and the solution is a uniformly
dark amber brown, remove the glass lid and parafilm, and very
gently slide each cassette into the trough of silver albumin stain
solution. The troughs that we use fit five cassettes laid flat along
the bottom of the trough. Do not stack the cassettes because
this will limit stain flow into each cassette. (If you have more
than four specimens/cassettes, then you will either need to
make additional troughs of stain (if you want to stain all speci-
mens the same days) or stain more specimens in the same
trough/stain solution starting the next day.) Try not to agitate
or otherwise disturb the stain solution. If cassettes need to be
moved around, then do this gently, without making waves or
splashes within the solution. When all the cassettes are in place
and fully immersed in stain solution, replace the parafilm and
glass lids, close the oven door, and leave specimens to stain
overnight. We typically start this process around 3:30 p.m. and
resume at 9:30 a.m. the following day (~18 h total incubation).
5. The next morning, gently remove the cassettes from the silver
stain solution. This solution can be kept in the oven for reuse or
new staining for up to 2 days, if not longer (we have never
needed to try it beyond 48 h). Tip each embryo back into its
original container and fill with dH2O. You will notice that the
dH2O will begin to yellow as some stain leaches out of the

embryo: this is normal and not a cause for concern. In fact, we
deliberately allow at least an hour for this leaching process to
remove “excess” stain. For longer-term storage, we suggest
using 100% ethanol which seems to limit the leaching of stain
beyond this initial “excess.” However, if specimens are being
scanned over a 48 h period, then it is fine to keep the tissues in
dH2O.
3.5 Embryo 1. Flip open the flat cap of a 1.5 mL clear, conical plastic micro-
Scanning centrifuge tube (see Note 4) and, using blunt forceps or a
similar tool, press in enough parafilm to completely fill the
hollow area of the lid. When the closed tube is turned upside
down (i.e., tapered end up), the parafilm-stuffed lid acts as a
“stage” on which the embryo can lie without touching any of
the walls of the tube. This step is not essential, but it’s useful
for minimizing the amount of plastic tube that the scanner’s
energy beam has to pass through to reach the specimen. Also,
keeping the embryo away from the tube walls makes it easier
to process and work with the raw scan data: if the embryo is
touching the solid plastic tube wall, then this will appear in the
images. If possible, it is even better to use dull forceps or some
other blunt microsurgery tool to gently push down into the
center of the parafilm and make a depression in which the
embryo can be nestled such that it sits upright (i.e., head up).
If the embryo is very small (e.g., E11), then we suggest
stuffing the tapered, bottom end of the conical tube with
parafilm, pressing a shallow indentation (or “nest”) into the
center of the parafilm and placing the bottom/caudal end of
the embryo into this “nest” so that the head sits above the
level of the parafilm. Ultimately, the trick is, first, to ensure
that the embryo will not move during the scan, which will of
course blur the images, and, second, to have as few layers
(e.g., plastic, parafilm) as possible between the beam of the
scanning system and the embryo to minimize bending or
other distortion of the beam before it reaches the specimen.
This same process can be done using a 7 mL screw-top tube; if
preferred, simply make a larger parafilm “nest” that fits in the
bottom of the tube.
2. For each embryo, we find it simplest to (1) suck off the ethanol
solution from the tube, and put the ethanol in a 50 mL tube for
short-term keeping and reuse once the specimen is scanned;
(2) tip the embryo from its original storage tube into the
parafilm-prepared scanning tube; and(3) into the tube pour
100% ethanol solution, which tends to help keep the specimen
in place better than water or PBS. There is no need to fill the
tube beyond the volume of solution required to just cover the
Fig. 3 The excitation of silver molecules (gray curve, top) suddenly rises at about
25.5 KeV, its K-edge, compared to water (orange line, bottom). This excitation is
the reason that the silver-impregnated tissues are detected and visualized with
particular clarity at a beam energy of 25.5 KeV
embryo. If using the parafilm “stage” in the cap/lid of the

tube, simply fill the upright tube (already containing the
embryo) with ethanol, close the tube, and flip it upside down.
Jiggle the tube to get the embryo to float into place on the
“stage.” Then the flat cap of the upside-down tube can sit flush
on the stage of the scanner. If scanning with the tube right-side
up (tapered end down), carefully pour ethanol into the tube,
gently position the embryo in its “nest,” and close the tube for
mounting in the scanner.
3. The small size of the embryos is well suited for high-
resolution imaging at the scale of tens to single microns
depending on the imaging system. The normally poor absorp-
tion contrast of the low-density embryos is greatly improved
by the impregnation of silver into the tissue. The optimal
energy for imaging silver lies just above 25.5 keV (see Note
8) (Fig. 3). For desktop systems, the energy spectrum of the
beam can be adjusted via the peak voltage (kVp) which sets the
maximal X-ray energy as well as filters which attenuate the
lower energies. For example, a system such as the SkyScan
1172 (Bruker Micro-CT) with a tungsten anode set at 40 kVp
and using 1 mm of aluminum for filtration has a mean X-ray
energy of 25.5 keV. A synchrotron source offers precise tun-
ability, enabling a beam with a narrow energy bandwidth
(a few hundred eV). Ultimately, image quality and resolution
will be a product of the system and scanning parameters used.
Once acquired, the serial 2D tomographic datasets can be
manipulated in either 2D or 3D by a variety of software
packages, depending on the scanning system and your soft-
ware preferences (see Note 9).
4 Notes
1. Thaw fixative in lukewarm water for at least 30 min in advance of

the time that you want to use this solution. Do NOT microwave
fixative solution, which somehow limits its efficacy: specimens
will not be properly fixed. The 4% PFA solution can be made up
at least 4 months in advance and stored at 20 C. Once fixative
is thawed, it can be stored at 4 C for about a week without
losing potency. Ideally, fixative is handled while working inside a
fume hood to protect the user from splashes and fumes. If using
a 5% glutaraldehyde/4% PFA solution, add the glutaraldehyde
once the PFA solution is fully thawed and is clear not cloudy and
no white precipitate is visible when the PFA is gently agitated.
Or, you can make the glutaraldehyde/PFA solution in advance,
and freeze this at 20 C for several months.
2. While we do not adhere to a hard and fast rule about the
volume of fixative or ethanol relative to embryo size, at the
very least, the container must be large enough to hold 10–20
times the volume of fluid compared to the embryo volume.
Further, the container must allow for solution to freely circu-
late around the embryo.
3. Most if not all synchrotrons hold calls or competitions for
allotted shifts of time on a “beamline,” which refers to the
particular scanning energy and capabilities of the light at that
particular “station” of the synchrotron. Each synchrotron is a
vast (e.g., football field-sized) microscope composed of a cen-
tral ring through which electrons are accelerated and then
routed to the various end stations, called beamlines. Potential
users typically need to contact the scientists working at the
particular beamline that the user wants to access for scanning
specimens. Each synchrotron has its own application process
and evaluation rubric, but typically the argument for “beam
time” rests on the unique need to scan specimens using syn-
chrotron light versus more conventional and accessible mod-
alities, such as a desktop μCT scanner.
4. Plastic is better than glass in terms of minimizing light/photon
diffusion and reflection.
5. If specimens are older than about E15, when skin begins to
develop, specimens can be cut down the midline to improve
stain penetration. However, the more mineralized a specimen
is, the less the stain will penetrate. Specimens aged postnatal
day 0 (birth) and older do not stain well [6].
6. All steps involved in working with live laboratory animals require
appropriate animal use and research ethics training according to
institutional policies. Users will need to be trained in methods of
humane euthanasia. For our work we used cervical dislocation,
particularly because many of our embryos are collected for gene

expression analyses and chemical inhalants, or injectable solu-
tions may alter expression profiles in tissues.
7. Silver staining will not work unless the tissues are fully dehy-
drated. We know this because we forgot one time to do this
dehydration step, and we put the fixed embryos straight into
silver stain, and the scan images were very poor with low tissue
contrast.
8. The X-ray absorption of materials generally declines with
increasing energy. There are, however, discontinuities in this
pattern where absorption increases as the energies required to
bind electrons are crossed. For silver, crossing the K-edge
energy—the energy binding its closest (K-shell) electron
(~25.5 keV)—the absorption jumps by a factor of nearly
6 [15] (Fig. 3). Thus, when tuning energies above this level,
silver-impregnated tissues have greatly increased absorption,
effectively painting a bull’s eye on them, which maximizes
tissue contrast and thus structural detail at very high resolution.
9. We do not detail the steps required to process, render and
refine, or analyze the volumetric scan image data simply
because these steps are linked to the scanning system and
local resources available to each user. To execute these steps,
the user will need to read and learn about their particular μCT
system as well as their computing resources including image
analysis software and data storage. If your institute has a core
imaging/scanning facility, then we highly recommend this as
the starting place to investigate hardware and software options
and training.
Acknowledgments
We are grateful to Karen Yuen and the College of Medicine Histol-

ogy Core Facility for the help and resources required to optimize
this staining technique. We also thank College of Medicine Vivar-
ium staff, particularly Carmen Whitehead for maintaining JCB’s
mouse colony. JCB and DMLC are each funded by a Discovery
Grant from the Natural Sciences and Engineering Research Coun-
cil of Canada (NSERC; JCB (#2011-402148, #2016-05177);
DMLC (#2014-05563) and by the Canadian Foundation for Inno-
vation (CFI infrastructure grants to JCB (#29037). This work was
also supported by the Canadian Institutes of Health Research
(CIHR)-THRUST Fellowship program. The method described in
this article was developed in part at the Canadian Light Source,
which is supported by NSERC, CIHR, the National Research
Council Canada, Province of Saskatchewan, Western Economic
Diversification Canada, and University of Saskatchewan.
References
1. Nanci A (2012) Ten cate’s oral histology: health. J Exp Zool B Mol Dev Evol 328
development, structure and function, 8th edn. (4):321–333. https://doi.org/10.1002/jez.
Elsevier, Toronto b.22734
2. Berkovitz BKB, Holland GR, Moxham BJ 10. Radlanski RJ, Renz H, Zimmermann CA,
(2017) Oral anatomy, histology and embryol- Mey R, Matalova E (2015) Morphogenesis of
ogy, 5th edn. Elsevier, Toronto the compartmentalizing bone around the
3. Thesleff I, Tummers M (2009) Tooth organo- molar primordia in the mouse mandible during
genesis and regeneration. StemBook, The dental developmental stages between lamina,
Stem Cell Research Community, StemBook, bell-stage, and root formation (E13-P20).
doi:https://doi.org/10.3824/stembook.1. Ann Anat 200:1–14
37.1. http://www.stembook.org 11. Metscher BD (2009b) MicroCT for develop-
4. Lungova V, Radlanski RJ, Tucker AS, Renz H, mental biology: a versatile tool for high-
Misek I, Matalova E (2011) Tooth-bone mor- contrast 3D imaging at histological resolu-
phogenesis during postnatal stages of mouse tions. Dev Dyn 238:632–640
first molar development. J Anat 218:699–716 12. Silva JMS, Zanette I, Noël PB, Cardoso MB,
5. Chlastakova I, Lungova V, Wells K, Tucker AS, Kimm MA, Pfeiffer F (2015) Three-
Radlanski RJ, Misek I, Matalova E (2011) dimensional non-destructive soft-tissue visuali-
Morphogenesis and bone integration of the zation with X-ray staining micro-tomography.
mouse mandibular third molar. Eur J Oral Sci Sci Rep 5:14088. https://doi.org/10.1038/
119:265–274 srep14088
6. Raj MT, Prusinkiewicz M, Cooper DML, 13. Gignac PM, Kley NJ, Clarke JA, Colbert MW,
Belev G, Webb MA, Boughner JC (2014) Morhardt AC, Cerio D, Cost IN, Cox PG,
Imaging earliest tooth development in 3D Daza JD, Early CM et al (2016) Diffusible
using a new silver-based tissue contrast agent. iodine-based contrast-enhanced computed
Anat Rec 297:222–233 tomography (diceCT): an emerging tool for
7. Paradis MR, Raj MT, Boughner JC (2013) Jaw rapid, high-resolution, 3-D imaging of meta-
growth in the absence of teeth: the develop- zoan soft tissues. J Anat 228:889–909
mental morphology of edentulous mandibles 14. Schmidt EJ, Parsons TE, Jamniczky HA,
using the p63 mouse mutant. Evol Dev Gitelman J, Trpkov C, Boughner JC, Logan
15:268–279 CC, Sensen CW, Hallgrimsson B (2010)
8. Hammer CL, Franz-Odendall T (2016) What Micro-computed tomography-based pheno-
shapes the oral jaws? Accommodation of com- typic approaches in embryology: procedural
plex dentition correlates with premaxillary but artifacts on assessments of embryonic craniofa-
not mandibular shape. Mech Dev cial growth and development. BMC Dev Biol
141:100–108 10:18
9. Boughner JC (2017) Implications of verte- 15. http://physics.nist.gov/PhysRefData/
brate craniodental Evo-Devo for human oral XrayMassCoef/ElemTab/z47.html
Chapter 21
Isolation of SIBLING Proteins from Bone and Dentin Matrices

Abstract
The extracellular matrix of the bone and dentin contains several non-collagenous proteins (NCPs). One
category of NCPs is termed the SIBLING (small integrin-binding ligand, N-linked glycoprotein) family,
which includes osteopontin (OPN), bone sialoprotein (BSP), dentin matrix protein 1 (DMP1), dentin
sialophosphoprotein (DSPP), etc. These proteins have abundant phosphoserines, aspartic acids, and
glutamic acids. In this protocol, we describe the extraction of NCPs from the bone and dentin matrices
using guanidine-HCl/EDTA and the isolation of polyanionic SIBLINGs from NCPs using ion-exchange
fast protein liquid chromatography (FPLC) to separate the differentially charged proteins into different
fractions through a gradient elution by NaCl.
Key words Non-collagenous protein, Small integrin-binding ligand, N-linked glycoprotein, FPLC,
Ion-exchange chromatography, Ultrafiltration, Stains-all staining
1 Introduction
An early report described the method of isolating a phosphorylated

glycoprotein from a mixture of proteins extracted by demineraliza-
tion of rat bone with 0.5 M EDTA in 4 M guanidinium chloride
[1]. At that time, this phosphorylated glycoprotein (OPN), as well
as other SIBLING members, had not been identified or well char-
acterized. The chromatography method used for isolating OPN
actually wasted most of other SIBLING members during the pro-
cess. In this protocol, we employ a method of extracting NCPs
from bone/dentin matrix with 0.5 M EDTA/4 M guanidine chlo-
ride, followed by an ultrafiltration to change the buffer from 4 M
guanidine to 6 M urea. To isolate SIBLINGs from the NCP mix-
ture, we used ion-exchange FPLC to bond the polyanionic proteins
and elute them with a gradient concentration of NaCl. Stains-all
staining showed that the acidic SIBLINGs were enriched in differ-
ent elution fractions [2].
211
2 Materials
2.1 Extraction of 1. 10 mL syringe, 23-gauge needle.

Non-collagenous 2. Phosphate-buffered saline.
Proteins from the Bone
3. Mortar.
and Dentin
4. 4 M guanidine-HCl/inhibitor solution (Table 1) (see Note 1).
5. 4 M guanidine-HCl/0.5 M EDTA/inhibitor solution
(Table 2).
6. Platform rocker.
7. Falcon 50 mL conical centrifuge tubes.
8. 4 C refrigerator with glass door.
10. Polypropylene round bottom 50 mL ultracentrifuge tubes.
11. Beckman J2-21 centrifuge and JA-17 rotor.
2.2 Buffer 1. 6 M urea solution (4 L) (Table 3) (see Note 2).

Exchanging 2. Amicon stirred ultrafiltration cell (Fig. 1).
3. Compressed N2 cylinder (Fig. 1).
4. EMD Millipore Ultracel ultrafiltration disc (10 kDa) (Fig. 1).
5. Magnetic stirrer plate.
6. Deionized water.
Table 1
Composition of the 4 M guanidine-HCl/inhibitor solution
4 M guanidine-HCl/inhibitor solution
Guanidine-HCl 382.2 g (4 M)
Tris base (C4H11NO3) 6.05 g (50 mM)
EDTA (C10H12N2O8Na4·2H2O) 4.16 g (10 mM)
6-Aminohexanoic acid 13.12 g (0.1 M)
Benzamidine-HCl 0.78 g (5 mM)
Sodium iodoacetate 0.2 g (1 mM)
Soybean trypsin inhibitor 1.8 mg
Phenylmethylsulfonyl fluoride 0.17 g (10 mM)
Pepstatin solution (0.5 mg/mL) 10.0 mL (5 mg/L)
Deionized water To 1 L
SIBLING Isolation from Bone and Dentin 213
Table 2
Composition of the 4 M guanidine-HCl/0.5 M EDTA/inhibitor solution
4 M guanidine-HCl/0.5 M EDTA/inhibitor solution

Guanidine-HCl 382.2 g (4 M)
EDTA (C10H12N2O8Na4·2H2O) 208.1 g (0.5 M)
6-Aminohexanoic acid 13.12 g (0.1 M)
Benzamidine-HCl 0.78 g (5 mM)
Sodium iodoacetate 0.2 g (1 mM)
Soybean trypsin inhibitor 1.8 mg
Phenylmethylsulfonyl fluoride 0.17 g (10 mM)
Pepstatin solution (0.5 mg/mL) 10.0 mL (5 mg/L)
Table 3
Composition of 6 M urea solution (4 L)
6 M urea (pH 7.2)

Urea (NH2CONH2) 360 g (6 M)
2.3 Separation of 1. Buffer A: 6 M urea/0.1 M NaCl solution (Table 4).

SIBLING Proteins Using 2. Buffer B: 6 M urea/0.8 M NaCl solution (Table 5).
Fast Protein Liquid
3. Q Sepharose fast flow (GE).
Chromatography
(FPLC) 4. AKTA FPLC biomolecule purification system (GE).
5. XK 16/20 column (GE).
6. Disposable glass tubes.
2.4 Stains-All 1. Stains-all staining solution (Table 6) (see Note 3).

Staining 2. Fixation solution (Table 7).
3. Washing solution (Table 8).
4. Destaining solution (Table 9).
5. Vacuum gel drying system.
Fig. 1 Schematic diagram of the stirred ultrafiltration cell and experimental setup used for buffer exchange: (1)
permeate outlet, (2) ultrafiltration disc, (3) ultrafiltration cell, (4) stirrer, (5) blow-off valve, (6) nitrogen inlet
pressure, (7) valve, (8) manometer, (9) nitrogen, (10) magnetic stirrer plate, and (11) measuring cylinder.
Reproduced from ref. 3 with permission from Elsevier
Table 4
Composition of Buffer A: 6 M urea/0.1 M NaCl solution
6 M urea/0.1 M NaCl (pH 7.2)

Sodium chloride 5.84 g (0.1 M)
Table 5
Composition of Buffer B: 6 M urea/0.8 M NaCl solution
6 M urea (pH 7.2)

Sodium chloride 46.75 g (0.8 M)
Table 6
Composition of the stains-all staining solution
Stains-all 10 mg
Formamide 5 mL
Isopropanol 25 mL
3 M Tris, pH 8.8 0.5 mL
Deionized water To 100 mL
Table 7
Composition of the fixation solution
Methanol 45% (V/V)

Acetic acid 10% (V/V)
Deionized water 45% (V/V)
Table 8
Composition of the washing solution
Methanol 20% (V/V)

Table 9
Composition of the destaining solution
Methanol 45% (V/V)

3 Methods
3.1 Extraction of 1. Cut off the epiphyseal ends of long bones (see Note 4) to
Non-collagenous expose bone marrow. Flush away bone marrow with pre-cold
Proteins from the Bone PBS using a syringe and a 23-gauge needle.
and Teeth 2. Crush the bone or teeth in a pre-cold mortar containing
ice-cold 4 M guanidine-HCl/inhibitor solution. Remove any
soft tissues floating up in the solution, and transfer the crushed
tissues to a 50-mL falcon tube containing 40 mL pre-cold 4 M
guanidine-HCl/inhibitor solution. Shake the tube on a plat-
form rocker overnight at 4 C (see Note 5).
3. Centrifuge at 2500 g for 20 min at 4 C to spin down the

tissues. Collect the supernatant, and add 45 mL fresh ice-cold
4 M guanidine-HCl/0.5 M EDTA/inhibitor solution to
resuspend the pellet. Shake on the rocker for 5 days at 4 C
(see Note 6).
4. Centrifuge at 20,000 g for 20 min at 4 C to spin down the
tissues. Collect the supernatant and combine with the superna-
tant from step 3.
3.2 Buffer 1. Set up the ultrafiltration system (Fig. 1) in a 4 C refrigerator or

Exchanging cold room (see Note 7).
2. Transfer the crude extraction from Subheading 3.1 to a
pre-cold ultrafiltration cell, and add ice-cold 6 M urea solution
to the final volume of 400 mL.
3. Turn on the stirrer and N2 valve to apply 50 psi pressure on the
ultrafiltration cell. Put a beaker under the permeation outlet to
collect the permeated liquid waste.
4. When the sample volume in the ultrafiltration cell is condensed
into ~40 mL, turn off the N2 valve and stirrer. Turn on the
blow-off valve to release the pressure. Add 360 mL fresh
pre-cold 6 M urea solution to the condensed sample.
5. Repeat steps 3 and 4 three times.
6. Centrifuge at 20,000 g for 20 min at 4 C to spin down the
tissues. Collect the supernatant and store at 4 C.
3.3 Separation of 1. Set up Q Sepharose column: Fill a XK 16/20 column with a

SIBLING Proteins Using 10 mL bed volume of Q Sepharose fast flow. Allow the beads to
Fast Protein Liquid settle by gravity overnight.
Chromatography 2. Equilibrate the Q Sepharose column: Wash the column with
(FPLC) 50 mL deionized water and then 50 mL buffer A to equilibrate.
Flow, 2.0 mL/min (see Note 8).
3. Load 120 glass tubes to the fraction collector.
4. Separate SIBLINGs with FPLC: Load the injection tube with
50 mL crude extracts from Subheading 3.2 (see Note 9). Inject
49 mL extracts (see Note 10) into the Q Sepharose column;
then, apply elution program. Elution gradient: 0–88% buffer B
in 60 mL and 88% buffer B in 12 mL. Fraction: 1–120 (0.6 mL
each). Flow: 0.5 mL/min (Fig. 2).
5. Column sanitation: Wash the column with 20 mL 0.2 M
NaOH followed by 30 mL 20% ethanol.
3.4 Stains-All 1. Load 60 μL of each fraction from every three fractions onto a
Staining 4–20% gradient gel for SDS-PAGE analysis (please refer to
relevant chapter for the SDS-PAGE method).
2. Fix the PAGE gel in fixation solution for 1 h with gentle
agitation.
Fig. 2 Q Sepharose separation profiles of NCPs extracted from mouse’s long

bones. The NCPs were separated by anion-exchange chromatography through
gradient NaCl elution into 120 fractions, each containing 0.5 mL of 6 M urea
solution. Unpublished data by Dr. Xiaofang Wang
Fig. 3 Stains-all staining showed a smear pattern of NCP crude extracts from mouse’s long bones. Q
Sepharose anion-exchange chromatography separated and enriched SIBLING proteins into different fractions:
OPN was mainly eluted into fractions 42–48. BSP was primarily in fractions 57–81. The C-terminal fragments
of DMP1 were co-eluted with OPN and BSP in fractions 51–57. Reproduced from Ref. 2 with permission from
FASEB J
3. Wash the gel with 20% methanol for 1 h three times at room
temperature with gentle agitation.
4. Stain the gel in stains-all solution overnight at room tempera-
ture with gentle agitation (see Note 11).
5. Destain the gel in 45% methanol until the gel becomes clear
(Fig. 3).
6. For short-term storage, keep the gel in a dark box at 4 C.
7. For long-term storage, dry the gel with a vacuum gel drying
system according to the manufacturer’s instructions. Protect
the gel from light.
4 Notes
1. There are two types of EDTA: 4Na EDTA is easily dissolved in

water, while 2Na EDTA does not easily dissolve in water.
2. Add Tris first. Adjust the pH with HCl.
3. Protect the stains-all solution from light.
4. At least ten mice are needed in order to collect enough bone
tissues for NCP extraction. In addition to long bones, calvar-
ium is also a good source of bone matrix.
5. This step will extract NCPs from the unmineralized matrix.
6. This step will extract NCPs from the mineralized matrix.
7. Pre-wet the ultrafiltration disc in deionized water for 30 min.
8. If the column has been used before, wash the column with
50 mL buffer B to thoroughly remove the proteins previously
bonded to the column, followed by 50 mL buffer A to equili-
brate the column.
9. Avoid any bubbles in the injection tube and pipelines.
10. Only 49 mL samples will be loaded onto the chromatography
column. The remaining 1 mL crude extracts will be used as
controls in subsequent analyses.
11. Protect the staining from light.
References
1. Prince CW, Oosawa T, Butler WT et al (1987) 3. El-Abbassi A, Khayet M, Hafidi A (2011) Micel-
Isolation, characterization, and biosynthesis of a lar enhanced ultrafiltration process for the treat-
phosphorylated glycoprotein from rat bone. J ment of olive mill wastewater. Water Res
Biol Chem 262:2900–2907 45:4522–4530
2. Yang X, Yan W, Tian Y et al (2016) FAM20C is
the primary but not the only kinase for the SIB-
LINGs in bone. FASEB J 30:121–128
Chapter 22
Immunohistochemical Co-Localization of Amelogenin

and Ameloblastin in Developing Enamel Matrix
Rucha Arun Bapat and Janet Moradian-Oldak
Abstract
Quantitative co-localization analysis, combined with other in vivo and in vitro techniques, can provide
valuable information about the interaction and cooperative function of two proteins. Here we describe in
detail the technique of quantitative co-localization analysis of two enamel matrix proteins, amelogenin and
ameloblastin, in developing mouse enamel.
Key words Enamel, Enamel matrix proteins, Amelogenin, Ameloblastin, Immunolabeling, In vivo
quantitative co-localization, Manders’ coefficient
1 Introduction
Quantitative co-localization analysis of enamel matrix proteins was

first used in our laboratory by Gallon et al., in 2013 [1], to visualize
and quantify the co-localization of amelogenin and enamelin pro-
teins in vivo. Previous work had shown that amelogenin and enam-
elin interact in vitro [2, 3]. Co-localization between the two
proteins in developing mouse molars provided evidence that these
proteins also have cooperative function in vivo (Fig. 1). Hence,
combining co-localization analysis with in vitro functional analysis
was shown to be useful for elucidating protein-protein interactions.
We have further improved and applied this combination of
techniques to describe interaction between amelogenin and amelo-
blastin (Fig. 2) [4, 5]. We added fluorescence resonance energy
transfer (FRET) microscopy to the analysis to demonstrate close
proximity between two enamel extracellular matrix proteins
(Fig. 2c). FRET uses a pair of fluorophores such that the emission
wavelength of one is the excitation wavelength of the other (e.g.,
FITC and TRITC donor-acceptor pair). The efficiency of energy
transfer is inversely proportional to the sixth power of the distance
between them. The distance between two fluorophores has to be
219
220 Rucha Arun Bapat and Janet Moradian-Oldak
100
Amelogenin
Co-localization Coefficients (%)
90
Enamelin
80
70
60
50
40
30
20
10
0
B 0 10 20 30
Region of Interest
Fig. 1 Confocal image of postnatal day 1 mouse mandibular molar showing (a)
regions of interest (red circles) selected for quantitative co-localization analysis
along the secretory face of ameloblasts; (b) graph of co-localization coefficients
from ROIs shown in a. Inset—scatterplot showing distribution of red and green
pixels. Reproduced with permission from Ref. [1]
less than 10 nm to achieve a positive FRET signal. Using this

technique, Mazumder et al. [5] showed that amelogenin and ame-
loblastin proteins not only co-localize but are also within interact-
ing distance of one another.
The most commonly used measure of enamel matrix protein
co-localization is the Manders’ co-localization coefficient [6]
described later in this chapter. One of the key benefits of Manders’
co-localization coefficient over other coefficients (like Pearson’s
Co-Localization of Enamel Matrix Proteins 221
Fig. 2 Quantitative co-localization of ameloblastin and amelogenin in P8 mouse mandibular molar (a, b)
transverse section. Co-localization between amelogenin (green) and ameloblastin (red) is revealed by over-
lapping signals resulting in yellow staining. (b) Co-localization pattern of amelogenin and ameloblastin in
sagittal section. (d, e) White pixels exhibiting actual co-localization in the confocal images after background
correction within thickness of the enamel and around the rod sheaths. (c) Amelogenin-ameloblastin in situ
FRET efficiency displayed as an absolute range from highest (red 0.89) to lowest (purple 0.0). Small regions of
interest (ROIs) were chosen to obtain the FRET efficiency values around the rod sheaths (areas drawn with
white lines). Am ameloblasts, D dentin. Reproduced with permission from Ref. [5]
co-localization coefficient [7] or Manders’ overlap coefficient [6])

is that it calculates the fraction of each protein that co-localizes with
the other protein in a given area [8] irrespective of its intensity. This
is important, because the matrix proteins tagged by fluorophores
may not be evenly distributed within the enamel.
In this chapter, we will describe the technique of immunolabel-
ing developing mouse enamel for amelogenin and ameloblastin
proteins followed by co-localization analysis, using postnatal day
5 mouse mandibles as a model. This technique can be modified to
immunolabel postnatal day 8 teeth or other mineralized tissues.
2 Materials
2.1 For Paraffin 1. 4% paraformaldehyde (PFA): Heat 350 mL deionized (dI) water
Embedding of Mouse to 60 C. Add 20 g PFA and 50 mL 10 PBS (see below). Adjust
Mandibles pH to 7.4 with 5 N NaOH and bring final volume to 500 mL.
Filter using 0.45 μm sterile disposable filters (Nalgene™ Rapid-
Flow™), and store at 4 C (see Note 1).
2. 10% EDTA with 0.1% glutaraldehyde in PBS: Dissolve 50 g
disodium salt of EDTA, 50 mL 10 PBS, and 0.5 mL glutar-
aldehyde in 300 mL dI water. Adjust pH to 8.5 using 5 N
NaOH (see Note 2), and bring final volume to 500 mL. Filter
using 0.45 μm sterile disposable filters and store at 4 C.
3. 10 Phosphate-buffered saline (PBS): Dissolve 80 g NaCl, 2 g
KCl, 14.4 g Na2HPO4, and 2.4 g KH2PO4 in 800 mL dI
water. Adjust pH to 7.4 with 1 M NaOH or HCl, and bring

final volume to 1 L (see Note 3).
4. Ethanol: increasing concentrations from 25% to 100%.
5. Xylene.
6. Paraffin wax.
7. Vacuum-sealed hot air oven.
8. Plastic base molds and mounting rings.
9. Hot plate.
10. Forceps, etc. for handling tissue.
2.2 For 1. Citrate buffer: Dissolve 1.47 g sodium citrate dihydrate in

Immunohisto- 400 mL of dI water. Adjust pH to 6.0 with 1 M HCl and
chemical Labeling bring final volume to 500 mL. Add 0.05% Tween 20 to the
Coplin jar each time before use.
2. Tris-buffered saline (TBS): Dissolve 2.42 g Tris and 9.0 g NaCl
in 800 mL dI water. Adjust pH to 7.5 with 1 M HCl and bring
final volume to 1 L.
3. 0.3% Hydrogen peroxide: Dilute 0.1 mL of 30% H2O2 in
10 mL dI water.
4. 1% Bovine Serum Albumin (BSA): Dissolve 0.1 g BSA in 10
mL TBS. Store at 4 C.
5. Antibody dilution solution: 0.1% BSA, 0.3% Triton X-100
in TBS. Dissolve 0.01 g BSA and 30 μL Triton-X100 in 10
mL TBS. Store at 4 C.
6. Primary antibody: Anti-full-length amelogenin (host: chicken;
from Dr. Malcolm Snead’s lab), diluted 1:1000 in antibody
dilution solution.
7. Primary antibody: Anti-ameloblastin M-300 (host: rabbit;
commercially available from Santa Cruz Biotechnology Inc.,
Santa Cruz CA), 1:500 diluted in antibody dilution solution.
8. Secondary antibodies: Bovine anti-chicken conjugated with
FITC (Santa Cruz Biotechnology Inc., Santa Cruz CA) and
donkey anti-rabbit conjugated with Alexa 594 (Thermo Fisher
Scientific Inc., USA), each 1:100 in TBS.
9. Mounting media: Vectashield with DAPI (Vector Laboratories
Inc., MS).
10. Positively charged microscope slides (like Superfrost™ Plus,
Thermo Fisher Scientific Inc., USA) and coverslips.
3 Methods
3.1 Tissue In this protocol we use postnatal day 5 (P5) mouse mandibles for
Preparation the convenience of relatively quick demineralization and tissue
processing time. The incisor and molar enamel are co-labeled for
amelogenin and ameloblastin using FITC and Alexa 594, respec-
tively, visualized using confocal microscopy.
1. Fixation: Euthanize and dissect the heads of P5 mice following
the appropriate Institutional Animal Care and Use Committee
guidelines. Place the heads in 4% PFA immediately for 24–48 h
in a cold room (4 C) on a rocker. For proper fixation, about
15 times the volume of the fixative to tissue is preferred.
Remove the heads from PFA the following day, and dissect
out the mandibles carefully under a dissecting microscope
without damaging the structures of the developing teeth (see
Note 1).
2. Decalcification: Transfer the dissected mandibles to 10% EDTA
with 0.1% glutaraldehyde in PBS. For P5 mandibles, up to 5
days of decalcification at 4 C of decalcification is necessary.
The EDTA solution may be changed once every 48 h.
3. Washing: After decalcification, gently remove the tissue from
EDTA, and place in PBS for washing (see Note 4). Wash the
tissue in 1 PBS for 1–2 h with at least four changes of the
buffer at 4 C.
4. Dehydration: Dehydrate the tissue using increasing grades of
ethanol as below (see Note 5).
Concentration Changes Time

25% Two to three changes 30 min each
5. Clearing: Clear the ethanol from the tissue using two to three
changes of xylene, 30 min each (see Note 6).
6. Paraffin infiltration: After the third change of xylene, place the
tissue in 50% xylene-paraffin mix for 30 min in a 60 C vacuum-
sealed oven. Then continue with 100% paraffin for three
changes, 30 min each in a 60 C vacuum-sealed oven. The
vacuum will draw out the xylene and paraffin will replace
it. Tissue can be left in paraffin in the oven overnight.
7. Embedding: Select an appropriately sized tissue base mold for

your tissue. Place it on a 60 C hot plate and fill it halfway with
molten paraffin wax. Position your tissue in the desired orien-
tation. Remove from the hot plate and let the paraffin solidify
partially. Now place the mounting ring on the base mold, and
fill it with liquid paraffin all the way to the top. Allow to
completely set, preferably overnight. Store at 4 C (see Note 7).
8. Sectioning: Place paraffin blocks on ice 15–20 min prior to
cutting to make sectioning easier. Cut 5–7-μm-thick sections
using a microtome, and place them on positively charged
microscope glass slides. Allow to dry on a hot plate overnight.
Store prepared slides at 4 C (see Note 8).
3.2 Double Labeling 1. Antigen retrieval: The paraffin needs to be completely removed
for the antibodies to be able to attach to their respective
protein antigens. To remove paraffin, fill one Coplin jar with
citrate buffer with Tween 20. Place it in a water bath set at
60 C with two jars of dI water. Once the buffer is warm, place
the slides in the buffer and leave overnight. Next morning,
remove the slides from the buffer, and transfer to one of two
water-filled Coplin jars; wash by dipping each slide ten times.
Move the slides to the second water jar, and let them cool to
room temperature. Rinse the slides with TBS for 30 min (see
Note 9).
2. Blocking: Cover the exposed tissue with 0.3% hydrogen perox-
ide for 30 min. This helps in reducing the autofluorescence and
background from endogenous peroxidase in the tissue. Discard
the H2O2, blot excess from the edges of the slide, and wash the
slides with TBS for at least 30 min. Then cover the tissue with
1% BSA to block nonspecific binding of antibodies.
3. Primary antibody: Blot the slides using Kimwipes to remove
the BSA. For the co-labeling technique, both the primary
antibodies are prepared in the same solution (see Note 10).
Dilute anti-full-length amelogenin antibody and anti-M300
ameloblastin antibody 1:1000 and 1:500 in antibody dilution
solution. Prepare a moist chamber for overnight primary anti-
body incubation. Place the slides in the moist chamber, and
cover the tissues with about 150–200 μL primary antibody
solution (see Note 11). Incubate overnight at room tempera-
ture in the moist chamber.
4. Secondary antibody: Wash slides for 30 min in TBS after blot-
ting to remove the primary antibody solution. Secondary anti-
body selection is particularly important in case of double
labeling as none of the species should cross-react with each
other. We use bovine anti-chicken antibody conjugated with
FITC (green) for amelogenin and donkey anti-rabbit antibody
conjugated with Alexa 594 (red) for ameloblastin; however,

any other appropriate pair of species and fluorophores can be
used. Dilute the secondary antibodies 1:100 in TBS in the
dark. Incubate the slides with secondary antibodies in a dark
moist chamber for 3–4 h.
5. Nuclear stain and coverslip: It is crucial to note that every step
after this must be done in the dark. Cover the Coplin jars in
aluminum foil to keep the slides in the dark while washing.
Wash the slides in TBS in the dark for at least 30 min. Place a
drop of Vectashield with DAPI (Vector Laboratories Inc., Bur-
lingame CA) on the slides, and place an appropriately sized
coverslip to cover the tissue (see Note 12). Let the mounting
medium set for at least 30 min. Coverslips can be further sealed
using a clear nail varnish and dried for 15 min.
6. Visualization: The slides are visualized using a Leica SP8 con-
focal microscope with 63 oil immersion lens, or similar equip-
ment. Co-localization patterns within the developing incisor
are analyzed with Leica Application Suite (LAS) X version
1.8.1.13759 or equivalent.
3.3 Co-localization Before starting co-localization analysis, adjust the threshold and
Analysis background correction values. Keep these constant throughout the
entire analysis, across all the images/regions of interest (ROIs) to
be compared. Amelogenin and ameloblastin are known to be
secreted in different quantities, with amelogenin being more abun-
dant than ameloblastin. Therefore, it is important to take into
account the difference in the amount of each fluorophore that
will occur due to difference in the amount of the proteins. Man-
ders’ co-localization coefficient [6] factors in this difference in the
number of pixels of each channel in each selected ROI. Simply, it
calculates the ratio of the sum of co-localizing pixels of each chan-
nel to the total pixels in the ROI. Hence it generates two coeffi-
cients for two channels being considered for co-localization
(Fig. 3). Manders’ coefficients M1 and M2 are given by the follow-
ing equations [6]:
Fig. 3 Schematic diagram illustrating Manders’ co-localization coefficient. A, M1 ¼ 0.0, M2 ¼ 0.0; B,

M1 ¼ 0.250, M2 ¼ 0.250; C, M1 ¼ 0.500, M2 ¼ 0.500; D, M1 ¼ 0.500, M2 ¼ 0.400 [9]
P
i R i , coloc Sum of red pixels colocalizinng with green co localizing pixels in the ROI
M1 ¼ P ¼
R
i i Sum of total red pixels in the ROI
P
i Gi , coloc Sum of green pixels colocalizing with red co localizing pixels in the ROI
M2 ¼ P ¼
G
i i Sum of total green pixels in the ROI
where Ri,coloc ¼ Ri if Gi > 0 and Ri,coloc ¼ 0 if Gi ¼ 0, and
Gi,coloc ¼ Gi if Ri > 0 and Gi,coloc ¼ 0 if Ri ¼ 0 [6].
For amelogenin and ameloblastin co-localization, we selected
oval ROIs 3x5 μm in diameter within ameloblasts and at the
secretory face of ameloblasts. The Leica Application Suite software
provides Pearson’s correlation, overlap coefficient, co-localization
rate, co-localization area, ROI area, foreground area, background
area, mean intensity ROI, mean intensity co-localization, intensity
sum ROI, and intensity sum co-localization for each channel in
every ROI. Using these data, Manders’ co-localization coefficients
M1 and M2 can be calculated, and graphs can be plotted similar to
Fig. 1b (see Note 13).
4 Notes
1. All steps involving paraformaldehyde must be done carefully

under a chemical/fume hood with appropriate personal pro-
tective equipment, taking care not to inhale PFA fumes. Make
sure to cover the flask with aluminum foil while stirring PFA
solution if the stirring is done outside the hood.
2. EDTA will not dissolve until the pH of solution reaches 8.5.
Sometimes it may be necessary to use NaOH pellets instead of
5 N solution to raise the pH to 8.5. Stirring for several hours or
overnight at pH 8.5 may also be needed. We add glutaralde-
hyde to the EDTA solution to prevent protein loss during
demineralization.

3. 10 PBS can be autoclaved and kept at 4 C for long-term
storage. Dilute with dI water for use.
4. Demineralized tissue is very delicate. Take care not to hold the
tissue by the developing incisors or molars to maintain their
microstructure.
5. Tissue can be left in the second change of 70% ethanol
indefinitely.
6. Xylene is toxic and flammable. Work with xylene under a
chemical hood with all appropriate personal protective equip-
ment. Discard used xylene in designated chemical waste con-
tainers. Dehydration, clearing, and paraffin embedding steps
can be automated by using a spin tissue processor like Thermo
Scientific™ STP 120.
7. It takes some practice to be able to position the mandibles
correctly in paraffin without giving rise to any bubbles.
A convenient trick is to keep a spirit lamp at hand and warm the

tips of the instruments used to position the sample. This pre-
vents the wax from suddenly solidifying around the cold metal
tips of the instruments, causing bubbles as they are withdrawn.
It is a good idea to practice with some old/non-essential
samples before starting with actual experimental tissue.
8. Start with thicker sections initially as they are easier to handle.
Thickness of sections can be reduced to 5 μm with practice.
Take care while positioning the sections that the tips of incisors
and cusps of molars are not folded/torn.
9. Do not shock the tissue with varying temperatures of buffers. If
the buffers have been previously prepared and stored at 4 C,
make sure to bring them to room temperature before use.
10. It is important to test the antibodies for cross-reactivity before
co-labeling using Western blotting. This chapter describes the
antibodies that are tried and tested in our lab; however, you can
use any antibodies that work best for your protein of choice
and do not cross-react with each other.
11. The volume of primary antibody solution depends upon the
number of tissue sections on each slide; enough should be
prepared to cover all the sections. Make the desired amount,
and then calculate the appropriate quantity of antibody to be
diluted. For example, if you have two slides with four to five
sections each, you might need 200 μL antibody solution per
slide. Therefore, prepare 400 μL which will need 0.4 μL anti-
amelogenin and 0.8 μL anti-ameloblastin primary antibody.
12. Alternately, DAPI can be added with secondary antibody if it is
available as a solution. The coverslip can be mounted using
glycerol.
13. To make calculations easier, data generated by the Leica Appli-
cation Suite software are first exported in a statistical software
like Microsoft Excel. M1 and M2 are calculated by dividing
intensity sum co-localization with intensity sum ROI for each
channel. These figures can then be conveniently plotted in a
line graph to determine the co-localization of two channels. A
helpful resource to understand various co-localization coeffi-
cients can be found at the Scientific Volume Imaging
B.V. website https://svi.nl/ColocalizationCoefficients.
References
1. Gallon V, Chen L, Yang X, Moradian-Oldak J interaction between the 32kDa enamelin and
(2013) Localization and quantitative amelogenin. J Struct Biol 166(1):88–94
co-localization of enamelin with amelogenin. J 3. Yang X, Fan D, Mattew S, Moradian-Oldak J
Struct Biol 183(2):239–249 (2011) Amelogenin-enamelin association in
2. Fan D, Du C, Sun Z, Lakshminarayanan R, phosphate buffered saline. Eur J Oral Sci
Moradian-Oldak J (2009) In vitro study on the 119:351–356
4. Mazumder P, Prajapati S, Lokappa SB, Gallon V, heredity, and panmixia. Philos Trans Royal Soc
Moradian-Oldak J (2014) Analysis of London Series A 187:253–318
co-assembly and co-localization of ameloblastin 8. Dunn KW, Kamocka MM, McDonald JH
and amelogenin. Front Physiol 5:274 (2011) A practical guide to evaluating colocali-
5. Mazumder P, Prajapati S, Bapat R, Moradian- zation in biological microscopy. Am J Physiol
Oldak J (2016) Amelogenin-ameloblastin spatial Cell Physiol 300(4):C723–C742
interaction around maturing enamel rods. J 9. Huygens Support Wiki: Scientific Volume Imag-
Dent Res 95(9):1042–1048 ing (Hilverum, The Netherlands) [The SVI-wiki
6. Manders E, Verbeek F, Aten J (1993) Measure- is a rapidly expanding public knowledge resource
ment of co-localization of objects in dual-colour on 3D microscopy, image restoration (deconvo-
confocal images. J Microsc 169(3):375–382 lution), visualization and analysis. Based on the
7. Pearson K (1896) Mathematical contributions wiki principle, it is open to contributions from
to the theory of evolution. III. Regression, registered visitors]. https://svi.nl/
ColocalizationCoefficients
Chapter 23
The Expression and Purification of Recombinant

Mouse Ameloblastin in E. coli
Jingtan Su, Rucha Arun Bapat, and Janet Moradian-Oldak
Abstract
Ameloblastin is the second most abundant enamel matrix protein, and is thought to be essential for
ameloblast cell polarization, cell adhesion, and enamel mineralization. However, studies of ameloblastin’s
function and its molecular mechanism have been limited due to difficulty in obtaining recombinant
ameloblastin in vitro. Here, we present a protocol for successful ameloblastin expression and purification
in E. coli.
Key words Ameloblastin, Enamel, Matrix protein, Expression and purification
Abbreviations
AMBN Ameloblastin
DMSO Dimethyl sulfoxide
IPTG Isopropyl β-D-1-thiogalactopyranoside
LB broth Luria-Bertani broth
PMSF Phenylmethylsulfonyl fluoride
His-tag An amino acid motif in proteins that consists of at least six histidine (His) residues,
6 aa
S-tag An oligopeptide derived from pancreatic ribonuclease A, 15 aa
Trx-tag Thioredoxin protein, 105 aa
1 Introduction
Enamel extracellular matrix is composed of proteins, including

amelogenin, ameloblastin, enamelin, and amelotin, and proteinases
matrix metalloproteinase-20 (MMP-20 or enamelysin) and
kallikrein-4 (KLK-4) [1]. Ameloblastin (AMBN), which is encoded
by a secretory calcium-binding phosphoprotein (SCPP) gene
located on chromosome 4q21 [2], is the second most abundant
229
230 Jingtan Su et al.
enamel matrix protein, accounting for roughly 5% of the matrix [3].

It is a two-domain, intrinsically disordered protein with one specific
and several non-specific calcium-binding regions [4].
Ameloblastin is crucial for enamel mineralization.
Ameloblastin-mutant mice develop severe enamel hypoplasia [5],
and mice with overexpression of ameloblastin exhibit imperfections
in their enamel that are evident on the nanoscale [6]. It has also
been reported recently that deletion of ameloblastin exon 6 is
associated with amelogenesis imperfecta [7]. However, the exact
function of ameloblastin and its molecular mechanism is still
unclear. It has been suggested that ameloblastin is a cell adhesion
molecule which adheres ameloblasts to the enamel extracellular
matrix [5], that it may interact with calcium ions [4, 8], that it
could act as a signal molecule [9], that serine phosphorylation of
ameloblastin is important for enamel formation [10], and that the
self-assembly of ameloblastin is crucial for the organization of
enamel extracellular matrix and formation of properly structured
enamel [11].
Ameloblastin is highly expressed by ameloblasts during the
secretory stage of amelogenesis [8]. It has several identified or
putative phosphorylation, O-glycosylation, and hydroxylation sites
[4, 12]. Soon after its secretion, ameloblastin is cleaved by matrix
metalloproteinase 20 [13]. The N-terminal cleavage products of
ameloblastin are stable and accumulate in the enamel prism sheaths,
while the C-terminal cleavage products are successively cleaved into
smaller peptides and ultimately lost [8, 14]. Full-length ameloblas-
tin is only found adjacent to the secretory face of Tomes’ process.
Thus, intact ameloblastin is a trace component of developing
enamel and has never been isolated in vivo. Mouse ameloblastin
has been successfully expressed in Drosophila melanogaster expres-
sion system using Schneider 2 cells in DES system [15]. However,
this is a slow and low-yielding process. For these reasons, we devel-
oped a technique to express mouse ameloblastin (GenBank
No. AAB93765.1) with cleavable Thioredoxin, Histidine Trx-,
His-, and S-tags in E. coli. We describe here how this can be
achieved, after which the protein can be purified using nickel affinity
chromatography, and the tags can be cleaved by enterokinase
(NEB). As shown in Fig. 1, the final product of this protocol was
of sufficient quality for biochemical and biophysical experiments,
and mass spectra confirmed that this product was ameloblastin. The
protein expressed in E. coli lacks the posttransitional modifications
of glycosylation and phosphorylation. The advantages of the E. coli
expression system are the relative high yield and ease of expression
and purification. This recombinant ameloblastin is suitable for sec-
ondary and tertiary structural studies.
Recombinant Mouse Ameloblastin 231
A B
r
de BN
la
d AM 41490.0 Da
n d
ei ifi
e
ot r
pr pu
250 kDa
75 kDa
50 kDa
37 kDa
25 kDa
20 kDa
41400 41450 41500 41550 41600
mass
Fig. 1 SDS-PAGE and mass spectra of purified mouse AMBN. (a) SDS-PAGE showed that the purity of the
AMBN obtained by this technique was of sufficient quality for biochemical and biophysical experiments and
the apparent molecular weight was not higher than the theoretical value. (b) Mass spectra of the band around
50 kDa in SDS-PAGE showed that the exact molecular weight of the purified protein was close to the
theoretical value (41459.8 Da), suggesting the purified protein was AMBN
2 Materials
1. E. coli strain BL21(DE3) pLysS (Stratagene).

2. 500 mL 4 and 50 mL LB broth (1.25 g LB broth powder in
50 mL deionized water).
3. LB/NZCYM agar plates with ampicillin.
4. 100 mg/mL ampicillin 1 g sodium salt of ampicillin in 10 mL
dI water.
5. 1 M IPTG (2.383 g of IPTG in 10 mL dI water Sigma-
Aldrich).
6. 0.5 M EDTA in water, pH 8.5.
7. 0.1 M benzamidine HCl (156.62 mg benzamidine HCl in
10 mL deionized water).
8. 1 M PMSF in DMSO (1.74 g PMSF in 10 mL DMSO Sigma-
Aldrich).
9. Ni-NTA Agarose (Qiagen).
10. Imidazole (Sigma-Aldrich).
11. Phenomenex C4 High Performance Liquid Chromatography
(HPLC) column (10 250 mm, 5 μm).
12. 16,000 units/mL enterokinase (New England Biolabs).

13. 8 M urea (4.8 g urea in 10 mL dI water).
14. Buffers for affinity chromatography:
(a) Lysis buffer (binding buffer): pH 8.0, 50 mM NaH2PO4,
500 mM NaCl, 20 mM imidazole.
(b) Washing buffer: pH 7.2, 50 mM NaH2PO4, 500 mM
NaCl, 50 mM imidazole.
(c) Elution buffer: pH 7.2, 50 mM NaH2PO4, 500 mM
NaCl, 500 mM Imidazole.
15. Buffers for HPLC:
(a) Buffer A: 0.1% TFA in water, filtered with 0.45 um filter.
(b) Buffer B: 0.1% TFA, 80% acetonitrile in water, filtered
with 0.45 um filter.
3 Methods
3.1 Day 1 Make 1. Make LB agar plates with 1:1000 dilution of 100 mg/mL
Bacterial Culture ampicillin.
Plates 2. Prepare recombinant BL21 E coli. with pET-32a (Novagen)
plasmid inserted with mouse ameloblastin gene (GenBank No.
AAB93765.1) having thioredoxin, histidine, and S-tags using
standard methods of bacterial cloning.
3. Plate the recombinant E coli on ampicillin agar plates and
culture overnight at 37 C.
3.2 Day 2 Make 1. Prepare 2 L LB media in 4 L flasks, 500 mL in each flask.

Starter Culture 2. Prepare 50 mL LB media in a separate 250 mL flask.
3. Autoclave all media at 121 C and allow it to cool. Store these
flasks at 4 C until used.
4. Add 50 μL of 100 mg/mL ampicillin to the 50 mL media.
5. Inoculate the 50 mL LB media supplemented with ampicillin
with a single colony from the BL21 agar plate. Seal and save the
plate for later use at 4 C.
6. Incubate 50 mL culture overnight in a shaking incubator at
37 C. This is the starter culture.
3.3 Day 3 Protein 1. Remove the starter culture from the shaker-incubator and keep
Expression at 4 C until used.
2. Add 500 μL of 100 mg/mL ampicillin to each flask of 500 mL
LB media.
3. Measure optical density (OD) of the culture using a UV-Vis
spectrophotometer at 595 nm. This will serve as the baseline or
“blank” measurement. Keep this for later reading.
4. Inoculate each 500 mL LB broth supplemented with ampicillin

with one fourth of the starter culture (day 2 step 6).
5. Take OD reading at 595 nm immediately after inoculation.
Further readings should be taken periodically to keep track of
the growth. Read the blank each time you measure the OD.
6. Induce each flask with 500 μL 1 M IPTG mentioned in materi-
als, no need to say again when the bacterial growth reaches
~0.75–0.8 OD at 595 nm.
7. Continue to take OD readings periodically. Bacterial growth is
usually slow right after induction, so it is best to check OD 2 h
after induction. Harvest the bacteria after 4 h.
8. Pour the contents of the flask into centrifuge bottles. Do not
overfill the bottles. Balance the weight and centrifuge for 6 min
at 8000 rpm (9700 g) at 4 C.
9. Discard the supernatant and keep the bacterial pellets at
20 C overnight.
3.4 Day 4 Protein All steps from this point forward should be done on ice or in a cold
Purification room (see Note 1).
1. Resuspend the bacterial pellets in lysis buffer (20 mL lysis
buffer/500 mL culture pellet).
2. Add 1 mM EDTA (400 μL of 0.5 M EDTA solution for 80 mL
lysis buffer), 1 mM benzamidine, and 1 mM PMSF (200 μL
each) (see Note 2).
3. Sonicate the bacteria for 30 min at amplitude 20%, 1 s on 1 s
off, with a precooled sonicator tip using an ultrasonicator like
Branson Sonifier (Branson Ultrasonics, US).
4. Centrifuge twice at 12,000 rpm (20,000 g) for 15 min at
4 C.
5. Prepare Ni-NTA columns by washing with 15 mL elution
buffer, followed by 20 mL lysis buffer.
6. Decant supernatant in clean Ni-NTA agarose gel tubes, and
place on a rocker for the proteins to bind for 1 h at 4 C (see
Note 3).
7. Let the supernatant pass through the columns. Add 50 mL
binding buffer (25 mL twice) and let it drip through the
Ni-NTA tube.
8. Add 50 mL washing buffer and let it pass through the Ni-NTA
tubes.
9. Elute proteins using 5 mL elution buffer and prepare for dialy-
sis as described below.
3.5 Days 4 and 5 1. Make 3 L dialysis buffer by diluting 30 mL pH 7.4 1 M

Dialysis NaH2PO4 in 2.970 L water, and cool it at 4 C.
2. Use 1 L dialysis buffer and add 1 mL 1 M PMSF (final concen-
tration 0.1 mM), 1 mL 0.1 M benzamidine (final concentra-
tion 0.01 mM), and 1 mL 0.5 M EDTA (final concentration
0.5 mM) and stir.
3. Take the eluted protein solution, and carefully place it in a
10,000 kDa dialysis membrane clamped at one end. Clamp
the other end and suspend in the dialysis buffer stirring over-
night. Change the buffer twice: once in the morning and then
again 4 h later. Collect the dialyzed protein 4 h after the second
change.
4. Determine the concentration of protein after dialysis using a
Pierce BCA or other appropriate method kit.
3.6 Day 6 Cleave the 1. At this point the protein is in pH 7.4 10 mM NaH2PO4 buffer
Trx-, His-, and S-Tags containing EDTA, PMSF, and benzamidine.
2. Add 8 M urea to the protein solution such that the final
concentration of urea is 1 M.
3. Enterokinase is used to cleave the tags to obtain Ameloblastin
in its native state. For 1 mg protein, 0.8 μL enterokinase
(12.8 units) is added. Calculate the amount of enterokinase
needed based on the concentration of protein after dialysis and
incubate the protein solution with enterokinase and urea at
37 C for 6 h with gentle mixing. Stop the reaction by adding
10% of the total volume of acetic acid, and store at 20 C
overnight.
3.7 Day 7 Remove 1. To separate the fragments of cleaved tags from full-length
the Cleaved Tags ameloblastin, HPLC (Varian) system is used. The system is
Using HPLC prepared by removing bubbles following the standard
protocol.
2. Clean the C4 column Phenomenex (10 250 mm, 5 μm)
following the standard column cleaning protocol.
3. Centrifuge the cleaved ameloblastin and keep the supernatant
to remove any solid particles. The amount of ameloblastin
injected in the HPLC column depends upon to the maximum
sample loop volume.
4. Elute with a gradient increasing from 40 to 90% buffer B for
80 min, at a flow rate of 1.5 mL/min.
5. Collect all the peaks. The first peak appears around 17 min.
There will typically be a four-peak pattern in which the second
peak is ameloblastin at around 30 min (see Note 4).
6. Lyophilize the collected peaks and dissolve them in distilled

water to run an SDS-PAGE.
7. Final protein should appear at around 50 kDa in 12%
SDS-PAGE gel.
4 Notes
1. Ameloblastin is not stable at room temperature, particularly

when the purity is low. For this reason, do not keep the protein
or protein mixture at room temperature.
2. Use freshly prepared benzamidine and PMSF stock can be
stored for several weeks at 20 C.
3. Ni-NTA can be reused up to 5 times. Follow manufacturer’s
recommendations for cleaning and regenerating the Ni-NTA
resin.
4. Ameloblastin is intrinsically disordered, and its aggregation
behavior is sensitive to the buffer conditions. The position of
peaks for ameloblastin in high-performance liquid chromatog-
raphy may therefore change slightly. It is advisable to run
SDS-PAGE to confirm the final product each time.
References
1. Moradian-Oldak J (2012) Protein-mediated 6. Paine ML, Wang HJ, Luo W, Krebsbach PH,
enamel mineralization. Front Biosci 17:1996 Snead ML (2003) A transgenic animal model
2. MacDougall M, DuPont BR, Simmons D, resembling amelogenesis imperfecta related to
Reus B, Krebsbach P, Karrman C, ameloblastin overexpression. J Biol Chem 278
Holmgren G, Leach RJ, Forsman K (1997) (21):19447–19452. https://doi.org/10.
Ameloblastin gene (AMBN) maps within the 1074/jbc.M300445200
critical region for autosomal dominant amelo- 7. Poulter JA, Murillo G, Brookes SJ, Smith CE,
genesis imperfecta at chromosome 4q21. Parry DA, Silva S, Kirkham J, Inglehearn CF,
Genomics 41(1):115–118. https://doi.org/ Mighell AJ (2014) Deletion of ameloblastin
10.1006/geno.1997.4643 exon 6 is associated with amelogenesis imper-
3. Krebsbach PH, Lee SK, Matsuki Y, Kozak CA, fecta. Hum Mol Genet 23(20):5317–5324
Yamada KM, Yamada Y (1996) Full-length 8. Murakami C, Dohi N, Fukae M, Tanabe T,
sequence, localization, and chromosomal Yamakoshi Y, Wakida K, Satoda T,
mapping of ameloblastin a novel tooth-specific Takahashi O, Shimizu M, Ryu O (1997) Immu-
gene. J Biol Chem 271(8):4431–4435 nochemical and immunohistochemical study of
4. Vymětal J, Slabý I, Spahr A, Vondrášek J, Lyng- the 27-and 29-kDa calcium-binding proteins
stadaas SP (2008) Bioinformatic analysis and and related proteins in the porcine tooth germ.
molecular modelling of human ameloblastin Histochem Cell Biol 107(6):485–494
suggest a two-domain intrinsically unstruc- 9. Zeichner-David M, Chen LS, Hsu Z, Reyna J,
tured calcium-binding protein. Eur J Oral Sci Caton J, Bringas P (2006) Amelogenin and
116(2):124–134 ameloblastin show growth-factor like activity
5. Fukumoto S, Kiba T, Hall B, Iehara N, in periodontal ligament cells. Eur J Oral Sci
Nakamura T, Longenecker G, Krebsbach PH, 114(Suppl 1):244–253.; discussion 254–246,
Nanci A, Kulkarni AB, Yamada Y (2004) Ame- 381–242. https://doi.org/10.1111/j.1600-
loblastin is a cell adhesion molecule required 0722.2006.00322.x
for maintaining the differentiation state of 10. Ma P, Yan W, Tian Y, He J, Brookes SJ, Wang X
ameloblasts. J Cell Biol 167(5):973–983 (2016) The importance of serine
phosphorylation of ameloblastin on enamel 13. Chun YH, Yamakoshi Y, Yamakoshi F,

formation. J Dent Res 95(12):1408–1414. Fukae M, Hu JC, Bartlett JD, Simmer JP
https://doi.org/10.1177/ (2010) Cleavage site specificity of MMP-20
0022034516661513 for secretory-stage ameloblastin. J Dent Res
11. Wald T, Spoutil F, Osickova A, 89(8):785–790. https://doi.org/10.1177/
Prochazkova M, Benada O, Kasparek P, 0022034510366903
Bumba L, Klein OD, Sedlacek R, Sebo P, 14. Uchida T, Murakami C, Wakida K, Satoda T,
Prochazka J, Osicka R (2017) Intrinsically dis- Dohi N, Takahashi O (1997) Synthesis, secre-
ordered proteins drive enamel formation via an tion, degradation, and fate of ameloblastin dur-
evolutionarily conserved self-assembly motif. ing the matrix formation stage of the rat incisor
Proc Natl Acad Sci U S A 114(9): as shown by immunocytochemistry and immu-
E1641–E1650. https://doi.org/10.1073/ nochemistry using region-specific antibodies. J
pnas.1615334114 Histochem Cytochem 45(10):1329–1340
12. Delsuc F, Gasse B, Sire JY (2015) Evolutionary 15. Ravindranath HH, Chen LS, Zeichner-David-
analysis of selective constraints identifies ame- M, Ishima R, Ravindranath RM (2004) Inter-
loblastin (AMBN) as a potential candidate for action between the enamel matrix proteins
amelogenesis imperfecta. BMC Evol Biol amelogenin and ameloblastin. Biochem Bio-
15:148. https://doi.org/10.1186/s12862- phys Res Commun 323(3):1075–1083.
015-0431-0 https://doi.org/10.1016/j.bbrc.2004.08.
207
Part IV
Protocols for Biochemistry and Imaging

Chapter 24
Protocols for Studying Formation and Mineralization

of Dental Tissues In Vivo: Extraction Protocol for Isolating
Dentin Matrix Proteins from Developing Teeth
Yasuo Yamakoshi, Jan C.-C. Hu, Mari M. Saito, and James P. Simmer
Abstract
The organic material in developing dentin is 90% type I collagen and 10% non-collagenous proteins. The
key to understanding dentin biomineralization is to study how these proteins collectively precipitate and
organize hydroxyapatite crystals. The first step in characterizing the proteins within a mineralizing matrix is
to efficiently extract and isolate the essential molecular participants and elucidate their structural and
biochemical properties. In this study, we expanded previous approaches to develop an improved strategy
for the extraction of extracellular matrix proteins from the dentin of developing teeth. Proteins in dentin
powder were sequentially extracted in the order Tris-guanidine buffer, HCl-formic acid solution, acetic
acid-NaCl solution, Tris-NaCl buffer, and a second Tris-guanidine buffer. Individual fractions were
analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), by gelatin or casein
zymography, and by Western blot analysis using dentin sialoprotein (DSP)- or dentin glycoprotein (DGP)-
specific antibodies. This approach was used to purify assorted porcine dentin non-collagenous proteins.
Key words Dentin, Teeth, Non-collagenous protein, Extraction, DSPP, Phosphophoryn
1 Introduction
Dentin is the mineralized tissue comprising the body of a tooth. It

protects the soft tissue pulp within and supports the overlying
enamel and cementum. On a weight basis, dentin is about 70%
mineral, 20% organic matrix, and 10% water. Dentin forms in a
defined extracellular space by matrix-mediated biomineralization.
Secreted proteins regulate and control the mineralization process.
With more than 40 years of experience isolating and characterizing
dentin proteins, many of the extraction and isolation procedures
have been optimized and standardized. Historically, a dissociative
extraction procedure using guanidine and the chelating agent
EDTA (ethylenediaminetetraacetic acid) was used for the extrac-
239
240 Yasuo Yamakoshi et al.
tion of proteins from dentin matrix [1–10], followed by

phosphoprotein-specific calcium precipitation [11], which effi-
ciently isolated dentin phosphoprotein (DPP), the highly acidic
C-terminal domain of dentin sialophosphoprotein (DSPP) [12].
In the early 1990s, the guanidine-EDTA-based method was also
used for the fractionation of dentin non-collagenous proteins other
than phosphoprotein such as dentin sialoprotein (DSP) [13–16].
We used this method to isolate and characterize DSP [17] and
dentin glycoprotein (DGP) [18] from developing porcine teeth.
Shortly thereafter, we developed a sequential extraction method
using successive extraction solutions, instead of the guanidine-
EDTA-based method [19]. The serial extraction method was later
improved using a rapid demineralization step, which replaced the
6-day acetic acid extraction with mixture of hydrochloric acid and
formic acid extraction for 1 day. The rapid demineralization step
greatly reduced postmortem proteolysis and allowed us to fraction-
ate intact dentin sialophosphoprotein (DSPP), the parent protein
of DSP, DGP, and DPP [20]. In this chapter, we explain the dentin
serial extraction protocol and subsequent methods for chro-
matographic separation of the major non-collagenous dentin
matrix proteins and visualize the results of these separations using
SDS-PAGE, zymography, and Western blot analyses.
2 Materials
2.1 Developing Tooth germs of permanent second molars were surgically extracted
Porcine Tooth using a hammer and chisel to remove them from the mandibles of
5-month-old pigs acquired from the Meat Market of the Metro-
politan Central Wholesale Market (Shinagawa, Tokyo, Japan)
(Fig. 1a). After removing the surrounding soft tissues and inner
pulp tissues with forceps, the tooth germs were washed in cold
saline and wiped carefully with a Kimwipe. The molars obtained
from these 5-month-old pigs were in the developmental stage of
advanced crown formation but prior to the onset of root formation
(Fig. 1b).
2.2 Extracting 1. 50 mM Tris–HCl/4 M guanidine (TG) buffer (pH 7.4):

Solutions for Weigh 6.057 g Tris and 382.12 g guanidine hydrochloride in
Sequential Extraction a glass beaker, and add purified water to a volume of 900 mL.
(See Notes 1 and 2) Mix and adjust pH with 6 N HCl to pH 7.4. Make up to 1 L
with purified water. Store at 4 C. Remove the contaminant in
buffer solution with a glass vacuum filter.
2. 0.17 N HCl and 0.95% formic acid (HF) solution: Add
42.4 mL of concentrated HCl and 80 mL of 95% formic acid
in a 10 L plastic pail. Make up to 8 L with distilled water. Store
at 4 C.
Sequential Extraction Protocol for Dentin Proteins 241
Fig. 1 Preparation of dentin powder from developing porcine tooth. (a) Tooth germs of permanent second
molars in the mandibles of a 5-month-old pig. (b) Surgically extracted second molar is in the crown formation
stage and has a mesial-distal dimension of about 2 cm. (c) The second molar after scraping off the enamel
layer. (d) 20 g of dentin powder in a container obtained by pulverizing 16 s molars
3. 0.5 M acidic acid/2 M NaCl (AN) solution: Weigh 116.88 g

NaCl in a 1 L graduated cylinder, and add purified water to a
volume of 900 mL. After the dissolution, add 29 mL of acetic
acid, and make up to 1 L with purified water. Store at 4 C.
4. 50 mM Tris–HCl/2 M NaCl (TN) buffer (pH 7.4): Weigh
6.057 g Tris and 116.88 g NaCl in a glass beaker, and add
purified water to a volume of 900 mL. Mix and adjust pH with
6 N HCl to pH 7.4. Make up to 1 L with purified water. Store
at 4 C. Remove the contaminant in buffer solution with a glass
vacuum filter.
2.3 Sodium Dodecyl 1. Novex 4–20% Tris-Glycine Mini Protein Gel (1.0 mm,
Sulfate- 12-well).
Polyacrylamide Gel 2. Novex 10% Zymogram (Gelatin) Protein Gel (1.0 mm,
Electrophoresis 10-well)
(SDS-PAGE) 3. Novex 12% Zymogram (Casein) Protein Gel (1.0 mm, 10-well)
2.3.1 SDS-PAGE and 4. NuPAGE 4–12% Bis-Tris Protein Gel (1.0 mm, 12-well)
Zymogram Gels (See
Note 3)
2.3.2 Running Buffer 1. Tris-Glycine SDS running buffer: Add 100 mL of Novex Tris-
(See Note 3) Glycine SDS running buffer (10) in a 1 L graduated cylinder,
and make up to 1 L with purified water. Store at room
temperature.
2. MES SDS running buffer: Add 50 mL of NuPAGE MES SDS
running buffer (20) in a 1 L graduated cylinder, and make up
to 1 L with purified water. Store at room temperature.
3. MOPS SDS running buffer: Add 50 mL of NuPAGE MOPS
SDS running buffer (20) in a 1 L graduated cylinder, and
make up to 1 L with purified water. Store at room temperature.
2.3.3 Sample Buffer (See 1. NuPAGE LDS Sample Buffer (4).

Note 3) 2. Novex Tris-Glycine SDS Sample Buffer (2).
2.3.4 Coomassie Brilliant 1. SimplyBlue SafeStain (Invitrogen-Thermo Fisher Scientific) for

Blue (CBB) Solution SDS-PAGE.
2. Coomassie Brilliant Blue (CBB) solution for Zymogram:
Weigh 2.5 g of CBB R-250 (BIO-RAD, Hercules, CA, USA)
in a glass beaker. Add 500 mL of methanol, 100 mL of acetic
acid, and 400 mL of purified water, and mix well. Filter into a
plastic reagent bottle. Store at room temperature.
2.3.5 Stains-All Solution 1. Stock solution: Weigh 0.1 g Stains-All (Sigma-Aldrich,

St. Louis, MO, USA) in 50 mL centrifuge tube, and dissolve
with 20 mL of formamide. Store at room temperature pro-
tected from light.
2. Staining solution: Mix 135 mL of 30 mM Tris–HCl buffer
(pH 8.8), 15 mL of formamide, 50 mL of methanol, and
1 mL of Stains-All stock solution in 200 mL graduated cylinder
(see Note 4).
2.3.6 Destaining Solution 1. Add 1.5 L of methanol, 300 mL of acetic acid, and 1.2 L of
for Zymogram purified water in a plastic reagent bottle.
2. Store at room temperature.
2.4 Western Blotting 1. Tris-Glycine transfer buffer: Add 40 mL of Novex Tris-Glycine

(See Note 5) transfer buffer (25) in a 1 L graduated cylinder, and make up
to 1 L with purified water. Store at room temperature.
2. NuPAGE transfer buffer: Add 50 mL of NuPAGE transfer
buffer (20) in a 1 L graduated cylinder, and make up to 1 L
with purified water. Store at room temperature.
3. Invitrogen PVDF/Filter Paper Sandwich (0.45 m pore size,
8.3 cm 7.3 cm).
4. Western Blotting Filter Paper, extra thick (20 cm 20 cm).
5. Tris-buffered saline (TBS)-Tween (TBST) solution: Add
200 mL of TBS (10) and 20 mL of 10% Tween 20 in a
graduated cylinder. Make up to 2 L with purified water. Store
at room temperature.
6. Blocking solution: Weigh 5 g of nonfat dry milk, and dissolve it
with 100 mL in TBST.
3 Methods
3.1 Preparation of 1. Scrape off the enamel layer with a curette from second molars
Dentin Powder (Fig. 1b).
2. Pulverize the remaining hard tissue to powder using a jaw
crusher (Fig. 1c) (see Note 6).
3. Yields approximately 20 g of tooth powder from 16 molars
(8 pigs) (Fig. 1d).
3.2 Sequential Carry out all procedures under ice-cold conditions unless otherwise
Extraction of Proteins specified.
from Tooth Powder
(Fig. 2)
3.2.1 First Tris- 1. Add 10 g of tooth powder to 250 mL for the plastic
Guanidine Extract centrifuge tube.
(G1 Extraction) (See 2. Suspend with 100 mL of TG buffer.
Note 7)
3. Homogenize using a Polytron homogenizer for 1 min at half
speed.
4. Centrifuge for 10 min at 15,900 g at 4 C.
5. Collect the supernatant (G1 extract) in a glass beaker.
6. Repeat steps 2–5 two more times for insoluble material.
7. Transfer G1 extract to a Spectra/Por 3 membrane.
8. Dialyze against 16 L of distilled water for 3 days at 4 C while
exchanging daily.
Dentin Powder
50mM Tris-HCl/4M guanidine (pH 7.4)
Sup Ppt
Dialysis Demineralization with 0.17N HCl/0.95% formic acid
Sup Ppt Sup Ppt

(G1S ext) (G1P ext) (HF ext)
0.5M acetic acid/2M NaCl
Sup Ppt
(AN ext) 50mM Tris-HCl/2M NaCl (pH7.4)
Sup Ppt
(TN ext)
50mM Tris-HCl/4M guanidine (pH 7.4)
Sup Ppt
(G2 ext) (RIS)
Dialysis
Sup Ppt
(G2S ext) (G2P ext)
Fig. 2 Flowchart showing the procedures used to produce the primary extracts for the characterization of
proteins from dentin powder. Sup supernatant, Ppt precipitate, ext extract, RIS residue insoluble materials
9. Separate the soluble (G1S extract) and insoluble (G1P extract)

fractions by centrifugation for 10 min at 15,900 g at 4 C.
10. Lyophilize both G1S and G1P extracts.
3.2.2 Demineralization 1. Pack the Tris-guanidine insoluble material in a dialysis bag

(HF Extraction) (See using a spatula from the plastic centrifuge tube as much as
Note 8) possible.
2. Suspend the remaining Tris-guanidine insoluble material in the
plastic centrifuge tube with 5 mL of HF solution.
3. Pipetting it several times and transfer to a dialysis bag (see
Note 9).
4. Repeat step 3 until the residue disappears as much as possible
(see Note 10).
5. Dialyze against 8 L of HF solution for 1 day at 4 C.
6. Transfer the dialysis bag contents to the centrifuge tube.
8. Collect the supernatant (HF extract) in an Erlenmeyer flask (see
Note 11).
9. Transfer HF extract to a Spectra/Por 3 membrane.
exchanging daily.
11. Lyophilize HF extract.
3.2.3 Acetic Acid-NaCl 1. Pack the HF insoluble material in the plastic centrifuge tube
Extract (AN Extraction) (See using a spatula from the surface of glass vacuum filter as much
Note 12) as possible.
2. Suspend with 100 mL of AN solution.
speed.
5. Collect the supernatant (AN extract) in the Erlenmeyer flask
(see Note 11).
7. Transfer AN extract to a Spectra/Por 3 membrane.
exchanging daily.
9. Lyophilize AN extract.
3.2.4 Tris-NaCl Extract 1. Pack the AN insoluble material in the plastic centrifuge tube
(TN Extraction) (See using a spatula from the surface of glass vacuum filter as much
Note 13) as possible.
2. Suspend with 100 mL of TN buffer.
speed.
5. Collect the supernatant (TN extract) in the Erlenmeyer flask
(see Note 11).
7. Transfer TN extract to a Spectra/Por 3 membrane.
exchanging daily.
9. Lyophilize TN extract.
3.2.5 Second Tris- 1. Pack the TN insoluble material in the plastic centrifuge tube
Guanidine Extract using a spatula from the surface of glass vacuum filter as much
(G2 Extraction) (See as possible.
Note 14) 2. Suspend with 100 mL of TG buffer.
speed.
5. Collect the supernatant (G2 extract) in the Erlenmeyer flask
(see Note 11).
7. Transfer G2 extract to a Spectra/Por 3 membrane.

exchanging daily.
9. Lyophilize G2 extract.
3.2.6 Preparation of 1. Pack the G2 insoluble material in the plastic centrifuge tube
Residue Insoluble Material using a spatula from the surface of glass vacuum filter as much
(RIM) (See Note 15) as possible.
2. Suspend with 100 mL of purified water.
4. Discard the supernatant.
6. Lyophilize insoluble material.
3.3 SDS-PAGE Our electrophoresis has been carried out using NuPAGE Bis-Tris
Mini Gels or Novex Pre-Cast Gels (Tris-Glycine SDS and Zymo-
gram Gels) and performed in accordance with the general informa-
tion and protocols for NuPAGE Bis-Tris Mini Gels or Novex
Pre-Cast Gels [21] (see Note 16). Here, we only show the prepara-
tion of dentin sample for SDS-PAGE and results of SDS-PAGE,
Western blot, and Zymogram (Fig. 3).
3.3.1 Sample 1. Weigh 300–400 μg of each lyophilized extract in an Eppendorf

Preparation tube, and add purified water as it becomes the concentration of
2 μg/μL.
2. Transfer 75 μL or 50 μL to another Eppendorf tube, and
dissolve with 25 μL of NuPAGE LDS Sample Buffer (4) or
50 μL of Novex Tris-Glycine SDS Sample Buffer (2).
3. Load 6–7 μL (9–10.5 μg) on the gels.
4. Run electrophoresis in accordance with the general informa-
tion and protocols.
3.3.2 CBB Staining 1. Rinse the gel with 100 mL of purified water for 5 min at three
times.
2. Stain the gel with 50 mL of SimplyBlue SafeStain for 1 h at
room temperature.
3. Wash the gel with 100 mL of purified water for at least 3 h.
3.3.3 Stains-All Staining 1. Rinse the gel with 200 mL of 25% methanol for 30 min at three
times.
2. Stain the gel with 200 mL of Stains-All solution overnight in
the dark.
3. Destain the gel with 200 mL of 25% methanol for at least 1 h.
A B C
G1S G1P HF AN TN G2S G2P RIS G1S G1P HF AN TN G2S G2P RIS G1S G1P HF AN TN G2S G2P RIS
kDa
250
148 6 6 6
2 9
98 7
64 8
50
36 1 3 3
10
22
4 4
16
5 5
6
D E
G1S G1P HF AN TN G2S G2P RIS G1S G1P HF AN TN G2S G2P RIS
kDa kDa
250 250
148 148
98
98 64
64 11 50
50 13
36
36 12
22
22
16
Fig. 3 Primary dentin extracts. The eight primary dentin extracts are G1S, G1P, HF, AN, TN, G2S, G2P, and RIS.
(a, b) Porcine dentin powder extracts analyzed by SDS-PAGE stained with CBB and Stains-All. (c) Western blot
of SDS-PAGE using the DSP antibody. (d, e) Gelatin and casein Zymogram of dentin extracts (see Note 22).
Number indicates protein bands which were identified in dentin extracts. 1 remaining enamel proteins, 2 acid
soluble collagen, 3 osteonectin (SPARC), 4 dentin glycoprotein (DGP), 5 osteocalcin (BGP), 6 high molecular
weight dentin sialoprotein (HMW-DSP), 7 dentin phosphoprotein (DPP), 8 albumin, 9 insoluble collagen, 10 low
molecular weight DSP (LMW-DSP), 11 matrix metalloprotease 2 (MMP-2), 12 kallikrein 4, 13 MMP-20
3.3.4 Zymography 1. Rinse the Zymogram Gel with 100 mL of 2.5% Triton X
100 for 30 min two times.
2. Incubate the gel with 200 mL of 50 mM Tris–HCl buffer
(pH 7.4) overnight.
3. Stain the gel with 100 mL of CBB for zymography at room
temperature for 30 min.
4. Destain the gel with 100 mL of purified water for at least 3 h
(see Note 17).
3.4 Western Blotting 1. Electrotransfer the gel onto the PVDF membrane which has
been carried out in accordance with protocols of overview for
Western blotting (Thermo Fisher Scientific) (see Note 18).
2. Block the membrane with 5% skim milk in TBST for at least 1 h
at room temperature (RT) (see Note 19).
3. Add primary antibody in TBST containing 5% bovine serum

albumin (BSA).
4. Incubate the membrane overnight in 4 C on a shaker.
5. Wash the membrane with TBST for 10 min three times on a
shaker at RT.
6. Add secondary antibody in TBST containing 5% BSA for 1 h
at RT.
7. Wash the membrane with TBST for 10 min three times on a
shaker at room temperature.
8. During the washing, prepare ECL solution following the pro-
portion of solutions A and B provided by the manufacture (see
Note 20).
9. Incubate the membrane for 5 min at RT.
10. Visualize the result in the dark room (see Note 21).
4 Notes
1. Except for the HF solution, prepare all solutions using ultra-

pure water (prepared by purifying deionized water), to attain a
resistance of 18 MΩ-cm at room temperature.
2. The TG, AN, and TN solutions contain 0.6 g benzamidine
(5 mM as final concentration), 1 mL of 1 M phenylmethylsul-
fonyl fluoride (PMSF) (1 mM as final concentration), and 1 mL
of 1 M 1,10-phenanthroline (1 mM as final concentration).
3. We purchased all SDS-PAGE and Zymogram Gels, running
buffers, and sample solutions from Invitrogen-Thermo Fisher
Scientific.
4. Prepare the staining solution the third end rinsed 5 min before.
5. We purchased all transfer buffers and filter paper from
Invitrogen-Thermo Fisher Scientific and TBS (10), 10%
Tween 20, and nonfat dry milk from BIO-RAD.
6. Mixer Mill MM400 (Retsch, Newtown, PA, USA).
7. The purpose of the G1 extraction is to remove remaining
enamel proteins (mostly amelogenins) and blood and tissue
components.
8. The purpose of the HF extraction is to demineralize and
thereby extract proteins having an affinity for hydroxyapatite.
9. Spectra/Por 3 Dialysis Membrane (nominal flat width, 54 mm;
diameter, 34 mm) (Spectrum Laboratories Inc., Rancho Dom-
inguez, CA, USA).
10. Final volume in dialysis bag should be less than 50 mL.
11. As the insoluble materials sometimes become smooth particles

after the demineralization, we mainly collect HF, AN, TN, and
G2 extracts through a glass vacuum filter.
12. The purpose of the AN extraction is to extract proteins ioni-
cally bound to collagen.
13. The purpose of the TN extraction is also to extract proteins
ionically bound to collagen and to neutralize for next
guanidine step.
14. The purpose of the G2 extraction is to extract proteins binding
to the collagen.
15. Because the main component in RIM is an insoluble collagen,
this step is optional.
16. Refer to: https://tools.thermofisher.com/content/sfs/
manuals/MAN0007891_NuPAGE_BisTris_MiniGels.pdf
17. We usually use five to six about the blocked cube sponge (each
size: 2 cm 2 cm 2 cm) to enhance the adsorption of the
CBB dye.
18. Refer to: https://www.thermofisher.com/jp/ja/home/
references/protocols/proteins-expression-isolation-and-analy
sis/western-blot-protocol/electrophoresis-and-blotting.
html#prot1
19. We sometimes perform the blocking step for overnight in 4 C
on a shaker.
20. We use Amersham ECL Prime Western Blotting Detection
Reagent (GE Healthcare Life Sciences, Uppsala, Sweden).
21. We used Amersham Imager 600 (GE Healthcare Life
Sciences).
22. The intact DSPP is extracted in G2P fraction, but it is not
detected by Western blot at this point because of a large
amount of collagen degradation products. We further fraction-
ate the intact DSPP by ion-exchange chromatography and
reverse-phase high-performance liquid chromatography.
Acknowledgment
This work was supported by the National Institute of Dental and

Craniofacial Research (NIDCR, US National Institutes of Health
[NIH]) grant DE018020, and JSPS KAKENHI, Grant-in-Aid for
Scientific Research (C; 26462982).
References
1. Veis A, Perry A (1967) The phosphoprotein of both dentin sialoprotein and dentin phospho-
the dentin matrix. Biochemistry 6 protein. J Biol Chem 273(16):9457–9464
(8):2409–2416 13. D’Souza RN, Bronckers AL, Happonen RP,
2. Veis A, Spector AR, Zamoscianyk H (1972) Doga DA, Farach-Carson MC, Butler WT
The isolation of an EDTA-soluble phospho- (1992) Developmental expression of a 53 KD
protein from mineralizing bovine dentin. Bio- dentin sialoprotein in rat tooth organs. J His-
chim Biophys Acta 257(2):404–413 tochem Cytochem 40(3):359–366
3. Dickson IR, Dimuzio MT, Volpin D, 14. Qin C, Brunn JC, Jones J, George A,
Ananthanarayanan S, Veis A (1975) The Ramachandran A, Gorski JP, Butler WT
extraction of phosphoproteins from bovine (2001) A comparative study of sialic acid-rich
dentin. Calcif Tissue Res 19(1):51–61 proteins in rat bone and dentin. Eur J Oral Sci
4. Butler WT, Hall WT, Richardson WS (1976) 109(2):133–141
Purification and some properties of the phos- 15. Qin C, Cook RG, Orkiszewski RS, Butler WT
phoprotein from rat incisors. Biochim Biophys (2001) Identification and characterization of
Acta 427(1):262–267 the carboxyl-terminal region of rat dentin sia-
5. Butler WT, Mikulski A, Urist MR, Bridges G, loprotein. J Biol Chem 276(2):904–909.
Uyeno S (1977) Noncollagenous proteins of a https://doi.org/10.1074/jbc.M006271200
rat dentin matrix possessing bone morphoge- 16. Qin C, Brunn JC, Baba O, Wygant JN, McIn-
netic activity. J Dent Res 56(3):228–232 tyre BW, Butler WT (2003) Dentin sialopro-
6. Butler WT, Bhown M, Dimuzio MT, Linde A tein isoforms: detection and characterization of
(1981) Noncollagenous proteins of dentin. a high molecular weight dentin sialoprotein.
Isolation and partial characterization of rat Eur J Oral Sci 111(3):235–242
dentin proteins and proteoglycans using a 17. Yamakoshi Y, Hu JC, Fukae M, Iwata T, Kim
three-step preparative method. Coll Relat Res JW, Zhang H, Simmer JP (2005) Porcine den-
1(2):187–199 tin sialoprotein is a proteoglycan with glycos-
7. Jontell M, Linde A (1977) Phosphoprotein of aminoglycan chains containing chondroitin
rat incisor dentine. Calcif Tissue Res 22 6-sulfate. J Biol Chem 280(2):1552–1560.
(Suppl):321–324 https://doi.org/10.1074/jbc.M409606200
8. Jontell M, Linde A, Lundvik L (1980) Com- 18. Yamakoshi Y, Hu JC, Fukae M, Zhang H, Sim-
parative studies of phosphoprotein prepara- mer JP (2005) Dentin glycoprotein: the pro-
tions from rat incisor dentin. Prep Biochem tein in the middle of the dentin
10(3):235–253 sialophosphoprotein chimera. J Biol Chem
9. Lee SL, Veis A, Glonek T (1977) Dentin phos- 280(17):17472–17479. https://doi.org/10.
phoprotein: an extracellular calcium-binding 1074/jbc.M413220200
protein. Biochemistry 16(13):2971–2979 19. Yamakoshi Y, Hu JC, Iwata T, Kobayashi K,
10. Munksgaard EC, Butler WT, WSd R (1977) Fukae M, Simmer JP (2006) Dentin sialopho-
Phosphoprotein from dentin. New approaches sphoprotein is processed by MMP-2 and
to achieve and assess purity. Prep Biochem 7 MMP-20 in vitro and in vivo. J Biol Chem
(5):321–331 281(50):38235–38243. https://doi.org/10.
1074/jbc.M607767200
11. Kuboki Y, Fujisawa R, Aoyama K, Sasaki S
(1979) Calcium-specific precipitation of dentin 20. Tsuchiya S, Simmer JP, Hu JC, Richardson AS,
phosphoprotein: a new method of purification Yamakoshi F, Yamakoshi Y (2011) Astacin pro-
and the significance for the mechanism of cal- teases cleave dentin sialophosphoprotein
cification. J Dent Res 58(9):1926–1932. (Dspp) to generate dentin phosphoprotein
https://doi.org/10.1177/ (Dpp). J Bone Miner Res 26(1):220–228.
00220345790580092001 https://doi.org/10.1002/jbmr.202
12. Feng JQ, Luan X, Wallace J, Jing D, 21. Penna A, Cahalan M (2007) Western Blotting
Ohshima T, Kulkarni AB, D’Souza RN, using the Invitrogen NuPage Novex Bis Tris
Kozak CA, MacDougall M (1998) Genomic minigels. J Vis Exp (7):264. https://doi.org/
organization, chromosomal mapping, and pro- 10.3791/264
moter analysis of the mouse dentin sialopho-
sphoprotein (Dspp) gene, which codes for
Chapter 25
Purification of Developing Enamel Matrix Proteins

Using Preparative SDS-PAGE
Steven J. Brookes and Claire M. Gabe
Abstract
In this chapter we discuss the potential of preparative SDS-PAGE for use in purifying native developing
enamel matrix proteins. We believe that the methodology has the potential to provide the relatively large-
scale single-step purification of any enamel protein that can be resolved as a single band during analytical
SDS-PAGE. Of course, a single band on analytical SDS-PAGE does not guarantee absolute purity as the
band may be comprised of two or more proteins migrating at the same apparent molecular weight on the
gel. Where absolute purity is required, the methodology can be used in conjunction with other techniques
such as ion-exchange chromatography or reverse-phase chromatography. We do not see preparative
SDS-PAGE replacing chromatographic methodologies but believe that it can provide another powerful
tool to add to the battery of purification techniques already available to researchers in the field.
Key words Preparative SDS-PAGE, Enamel matrix protein, Chromatographic methodologies
1 Introduction
The purification of proteins is an important research methodology.

Purified proteins have been used in a wide range of research studies
and are responsible for forwarding our general understanding of
the biological activity and function of specific proteins, protein
conformation, posttranslational modifications, and protein–protein
interactions. In addition, purified proteins have been used as immu-
nogens to generate antibodies for use as research tools. Purified
proteins have also been used as therapeutic agents. In the modern
era, protein purification has also played an essential role in the
recombinant protein revolution.
Not surprisingly, protein purification has played a significant
role in the development of the field of amelogenesis. Dental enamel
is incrementally secreted as a proteinaceous matrix by a monolayer
of ameloblast cells. The matrix begins to mineralize immediately
251
252 Steven J. Brookes and Claire M. Gabe
after secretion as the extremely elongated enamel hydroxyapatite

crystallites grow in length in pursuit of the retreating ameloblasts.
Analysis of the developing matrix proteins using protein purifica-
tion methodologies has contributed greatly to our present under-
standing about the nature of the enamel matrix. The major protein,
amelogenin, comprises >90% of the total protein while other minor
(but essential) components are derived from the enamelin and
ameloblastin genes. However, proteolytic processing of these pro-
teins by co-secreted matrix metalloprotease 20 generates a complex
mixture of polypeptides and peptides, a situation further compli-
cated by mRNA alternative splicing, potential sexual dimorphism,
and posttranslational phosphorylation of the dominant protein
amelogenin.
Early attempts to study developing enamel proteins, extracted
from the tissue using nondissociative gel filtration chromatography,
illustrated the concentration-dependent aggregation of enamel
matrix proteins and their propensity to exist as a multicomponent
system of large complexes [1]. These properties provided an early
indicator that their purification would be technically challenging.
Numerous gel filtration purification methodologies were published
in the following years that attempted to get round the problem of
aggregation by eluting columns with buffers containing chaotropic
agents or by using acidic or basic elution buffers. In addition, ion-
exchange chromatography was introduced as a further purification
step to further improve the degree of separation achieved (for
examples see [2–6]). The widespread introduction of reverse-
phase HPLC led to its application in purifying enamel matrix
proteins. Reverse-phase chromatography provided a step change
in the ability to obtain purified enamel matrix proteins [7–11].
Reverse-phase chromatography, used in conjunction with hydroxy-
apatite chromatography, ion-exchange chromatography, and affin-
ity chromatography (using the serine protease inhibitor
benzamidine as the immobilized ligand), was successfully used to
purify enamel matrix serine proteinase 1 (kallikrein 4), the protease
responsible for maturation-stage degradation of the enamel matrix
from maturing. The degree of purity obtained was sufficient for
amino acid sequencing of the protease [12], a significant achieve-
ment given that the protease is only present in the tissue at catalytic
concentrations.
Enamel protein purification based on chromatographic techni-
ques perhaps reached its zenith when gel filtration (which separates
proteins as a function of molecular size) was combined with chro-
matofocusing (which separates proteins based on their isoelectric
point) to achieve chromatographic separations analogous to those
achieved using 2D polyacrylamide gel electrophoresis [13].
Polyacrylamide gel electrophoresis in the presence of sodium
dodecyl sulfate (SDS-PAGE) has long been used as an analytical
technique that can separate proteins as a function of their molecular
Preparative SDS PAGE to Purify Enamel Matrix Protein 253
size with a high degree of resolution. The denaturing properties of

SDS make the technique well suited for analyzing many aggregative
or insoluble proteins as it readily solubilizes such proteins, and, in
the absence of any intramolecular covalent bonds, renders them
monomeric. Given its ability to cope with proteins with a propen-
sity to aggregate, SDS-PAGE is widely used in the literature to
analyze total enamel matrix proteins or subfractions of the matrix
generated during a purification procedure. However, such is the
propensity for amelogenin to aggregate, it can still dimerize to
some degree during SDS-PAGE [9].
In addition to its use in protein analysis, SDS-PAGE has also
been widely used in the general literature in a preparative manner
to fractionate proteins. However, preparative electrophoresis is
rarely reported in the amelogenesis literature even though it has
been successfully used in at least two occasions to further purify
enamel matrix proteins as an adjunct to chromatographic separa-
tion methodologies [14, 15]. In both these cases, samples were run
on standard analytical slab gels and the proteins visualized using
either trichloroacetic acid or Coomassie Blue, respectively. The
bands of interest were excised from the gels and the proteins
extracted from the bands using electroelution. This classic tech-
nique is simple to carry out without the need for any specialized
equipment but is limited by the amount of protein that can be
successfully resolved on a single slab gel. To get round this prob-
lem, companies have developed a range of commercial devices
designed to maximize the efficiency and applicability of preparative
PAGE. Such devices offer the potential to combine the high
resolving power of analytical PAGE with the sample-handling
capacity of lab-scale chromatography. We have already used prepar-
ative SDS-PAGE to purify amelogenin complexes stabilized in vivo
with molecular cross-linkers [16] and to purify recombinant ame-
logenin [17]. In this chapter, we discuss the potential of prepara-
tive SDS-PAGE for use in purifying native developing enamel
matrix proteins. We believe that the methodology has the potential
to provide the relatively large-scale single-step purification of any
enamel protein that can be resolved as a single band during analyt-
ical SDS-PAGE. Of course, a single band on analytical SDS-PAGE
does not guarantee absolute purity as the band may be comprised
of two or more proteins migrating at the same apparent molecular
weight on the gel. Where absolute purity is required, the method-
ology can be used in conjunction with other techniques such as
ion-exchange chromatography or reverse-phase chromatography.
We do not see preparative SDS-PAGE replacing chromatographic
methodologies but believe that it can provide another powerful
tool to add to the battery of purification techniques already avail-
able to researchers in the field.
2 Materials
2.1 Preparative PAGE The principle of preparative PAGE is no different to analytical

Apparatus: Design and PAGE; an electric field is applied to charged proteins to drive
Methods of Operation them through an acrylamide gel which acts as a molecular sieve.
The native charge on the protein may be augmented by the binding
of SDS, which gives proteins an approximately constant negative
charge-to-mass ratio so that separation is based on molecular size.
In the absence of SDS, the charge on the protein is simply the native
charge carried by the protein by virtue of any charged amino acid
side chains (or charged posttranslational modifications) associated
with the protein. In this case, separation is based on the charge-to-
mass ratio of the protein (which in turn depends on the pKa of the
constitutive ionizable groups and the environmental pH).
Commercial preparative PAGE systems can separate proteins
based on either separation modalities (SDS-PAGE or native charge
PAGE). There are four common methods for recovering proteins
separated using preparative PAGE: (i) As mentioned above, sepa-
rated proteins are harvested after electrophoresis on slab gels by
staining and excising bands of interest. The protein is then eluted
from the gel bands by electroelution or in some cases passive
diffusion. (ii) A more sophisticated approach is to electroelute the
whole gel which avoids the need to visualize the proteins. The
method is similar to Western blotting except that the eluted pro-
teins are not eluted onto a membrane but into an array of buffer-
filled channels or cells from which they can easily be recovered (e.g.,
Whole Gel Eluter as marketed by Bio-Rad). (iii) Separated proteins
are harvested during electrophoresis by continuously flowing col-
lection buffer over the end of a cylindrical gel to collect proteins as
they electrophorese off the bottom of the gel (e.g., Model 491 Prep
Cell and Mini Prep Cell by Bio-Rad). (iv) Two cylindrical gels are
separated by a space containing collection buffer. Periodically, elec-
trophoresis is paused and the collection buffer harvested. The space
is filled with fresh collection buffer and electrophoresis is continued
until the next batch of protein(s) is ready for collection (e.g., the
Nativen® system marketed by the ATTO Corporation).
The advantages and disadvantages of these systems have been
discussed previously [18] but we have adopted the continuous flow
system marketed by Bio-Rad because it provides a relatively high
sample capacity (Model 491 Prep Cell and Mini Prep Cell), and
good degree of resolution between the separated bands. The rest of
this chapter describes using the Bio-Rad Prep Cell to fractionate
porcine-developing enamel protein. It is not our intention to repli-
cate the manufacturer’s instruction booklet (which includes all the
details required to successfully optimize and carry out the method)
but rather to highlight and expand on the most critical aspects of
using the apparatus.
2.2 An Introduction Both the Model 491 Prep Cell and Mini Prep Cell employ a tube in
to the Model 491 Prep which the polyacrylamide gel is cast. The larger Model 491 Prep
Cell and Mini Prep Cell Cell is supplied with two gel tubes allowing gels to be cast with
diameters of 28 or 37 mm. The smaller Mini Prep Cell is provided
with a gel tube with a diameter of 7 mm. In both cases, proteins
migrate down the gel tube during electrophoresis and elute from
the bottom of the gel into an elution chamber through which
collection buffer (typically running buffer is used as the collection
buffer) is pumped to collect the proteins as they elute.
As shown in the simplified schematic in Fig. 1a the larger
Model 491 Prep Cell features a central ceramic cooling core
through which running buffer from the anode tank is circulated
which maintains a constant temperature gradient across the gel
during electrophoresis which ensures that proteins within a band
migrate at the same speed across the thickness of the gel. The
cooling circuit is shown in blue and a peristaltic pump is provided
to circulate the running buffer during electrophoresis. The collec-
tion buffer circuit is shown in red. Proteins eluting into the collec-
tion buffer are continuously drawn through the elution chamber
and up the center of the cooling core by a peristaltic pump provided
by the user. Proteins are then directed to a fraction collector
provided by the user. Collection buffer is stored in a reservoir that
surrounds the upper cathode buffer tank. The anode buffer present
in the larger lower buffer tank is in electrical contact with the
bottom of the gel tube through a reusable dialysis membrane
which prevents the eluted proteins from escaping from the elution
chamber into the anode buffer. We use the standard 6000 Da cutoff
dialysis membrane provided which is sufficient to retain even the
bromophenol blue tracking dye (molecular weight 670 Da) as the
collection buffer flowing through the elution chamber efficiently
removes molecules before they electrophorese through the dialysis
membrane into the anode buffer.
A simplified schematic of the Mini Prep Cell is shown in
Fig. 1b. The smaller diameter of the gel tube means that cooling
the gel during electrophoresis is not an issue and a cooling core is
not required. The collection buffer circuit shown in red is similar to
the system in Model 491 Prep Cell except that as proteins elute
from the bottom of the gel into the collection buffer they are
removed via a port in the base of the elution chamber and directed
toward a fraction collector.
Out of necessity, we have provided simplified schematics and
descriptions of both devices but the instruction manuals are freely
available from the manufacturer’s web site and contain exploded
diagrams of the devices and full details of their operation (see the
“Documents” section at http://www.bio-rad.com/en-ch/prod
uct/model-491-prep-cell-mini-prep-cell).
Fig. 1 (a) Simplified schematic diagram of the Model 491 Prep Cell. Blue lines
show the circulatory path of the anode buffer through the cooling core to maintain
an even transverse temperature differential across the gel during electrophoresis.
The red lines show the path of the collection buffer as it is drawn through the
elution chamber toward the fraction collector. As protein bands electrophorese
(elute) off the bottom of the gel, they are swept up the cooling core and to the
fraction collector by the flow of the collection buffer. A dialysis membrane
prevents eluted proteins from escaping into the anode buffer while allowing for
electrical contact between the gel and anode buffer. (b) Simplified schematic
diagram of the Mini Prep Cell. The smaller diameter of the gel tube means that
2.3 Protein-Loading The Model 491 Prep Cell is clearly designed to handle larger
Capacity protein loads than the Mini Prep Cell. According to the manufac-
turer’s specifications, the capacity of the larger Model 491 Prep Cell
is such that >4 mg of the target protein and its nearest contaminant
(not the total protein in the sample) can be loaded but this depends
on the degree of separation between the target protein band and its
nearest contaminant. An unavoidable feature of SDS-PAGE is that
protein bands widen as the amount of protein present in the band
increases. If a contaminating band is running at a similar molecular
weight then the loading limit will be reached when band broaden-
ing causes the bands to overlap. This can be overcome to some
extent by increasing the gel length to maximize band separation
during electrophoresis (as discussed in the section on optimizing
running conditions). According to the manufacturer, if the differ-
ence in molecular weight between the target protein and its nearest
contaminant approaches 2% then the maximum amount of target
protein that can be loaded may fall to <1 mg.
The manufacturer’s specifications for the Mini Prep Cell indi-
cate that the maximum amount of the protein of interest and its
nearest contaminant that can be loaded is <200 μg with the
amount falling to <50 μg if the molecular weight difference
approaches 2%.
3 Methods
3.1 Precautions and There are number of precautions that must be taken when using
Optimization of either prep cell to ensure a good separation. These are well detailed
Electrophoretic in the manufacturer’s instructions but we will mention the main
Conditions to Ensure factors that we have found to have the most influence on the
Good Separations separation obtained.
3.1.1 Casting the Gel Firstly, it is essential that the gel is cast correctly. For the Model
491 Prep Cell and the Mini Prep Cell the gel tubes are mounted on
a casting stand which incorporates a spirit level and adjustable feet
to make sure that the stand is flat and the gel tube is truly vertical
(see Note 1). As shown in Fig. 2, if these two surfaces are not
ä
Fig. 1 (continued) cooling is not required during electrophoresis. The red lines
show the path of the collection buffer as it is drawn through the elution chamber
toward the fraction collector. As protein bands electrophorese (elute) off the
bottom of the gel, they are swept out through a port in the base of the elution
chamber to the fraction collector by the flow of the collection buffer. A dialysis
membrane prevents eluted protein from escaping into the anode buffer while
maintaining electrical contact between the gel and the anode buffer. (Note, not
drawn to scale; the gel tube in the Model 491 Prep Cell is considerably bigger
than the gel tube in the Mini Prep Cell)
Fig. 2 Diagram showing the effect of casting gel when the gel tube is not perfectly vertical and perpendicular
to the pre-leveled casting base. (a) In the ideal case, the casting base has been leveled using the built-in spirit
level and adjustable feet. The upper and lower surfaces of the gel are parallel. The surface of the loaded
sample is parallel to the top and bottom gel surfaces and during electrophoresis bands migrate down the gel
parallel to the bottom of the gel and elute cleanly in the minimum time possible. (b) If the gel tube is not
vertical, the upper surface of the gel is not parallel to the lower surface of the gel. Although exaggerated here
to emphasize the point, the bands will electrophorese at an angle and will not run parallel to the lower surface
of the gel. This may lead to the co-elution of more than one band. As the band will take longer to fully elute, the
protein will be collected in a larger volume of collection buffer and will be unnecessarily diluted
Fig. 3 Diagram showing the possible consequences of polymerizing the resolving

gel with an air bubble trapped at the interface between the lower gel surface and
the casting stand. During collection of the target protein a contaminant following
closely behind can begin to partially co-elute with the target protein as it meets
the impression left in the gel by the air bubble
parallel then closely migrating bands will not be separated effi-

ciently as they will migrate down the gel at an angle (instead of
being perfectly parallel to the bottom surface of the gel) and will
co-elute to some degree.
Air bubbles trapped at the bottom of the gel can lead to an
uneven gel surface which may also cause co-elution of closely
migrating bands illustrated in Fig. 3 (see Note 2).
Heat generated during polymerization of the gel mixture can
cause inconsistencies in the gel. Any inconsistences in the gel will
cause the migrating bands to distort and they will no longer be
parallel to the bottom surface of the gel and the separation will
suffer (see Note 3).
Finally, enough time must be allowed for polymerization to be
complete so that inconsistences in the gel are minimized (see
Note 4).
3.1.2 Optimizing the Pore The pore size of the gel is determined by the concentration of
Size of the Gel (% acrylamide and bis-acrylamide monomers and must be selected so
Acrylamide and that the protein of particular interest is well resolved at the point it
Bis-Acrylamide Monomer is about to run off the bottom of the gel and enter the elution
Concentration) chamber. The manufacturer recommends that a total monomer
concentration is used that results in the target protein migrating
with a relative mobility of ~0.55–0.6 (see Note 5). The manufac-
turer provides a graph plotting recommended monomer concen-
trations against molecular weight of the target protein which users
can use to get a rough idea of what monomer concentration pro-

vides a relative mobility of ~0.55–0.6 (see Note 6). As detailed in
the manufacturer’s instructions, this is best achieved by using pre-
stained markers so that electrophoresis can be stopped when the
target protein of known molecular weight is approaching the bot-
tom of the gel. The gel is then stained and the distance between the
target proteins and its nearest contaminants is measured to see
which monomer concentration gives the best separation (see
Note 7).
3.1.3 Gel Height The target protein will elute more quickly from a shorter gel and
the protein in the band will be more concentrated due to reduced
band diffusion during the run and can be collected in smaller
volume of collection buffer (see Note 8). It is therefore a case of
determining the optimum gel height that will provide the required
degree of separation in the shortest run time without excessive
band diffusion.
3.1.4 Purging the System The elution chamber and associated tubes are purged with collec-
of Bubbles tion buffer prior to starting the run. Since SDS-PAGE running
buffer is used as the collection buffer, foaming and accumulation
of bubbles in the elution chamber and tubes can be a problem.
Degassing the collection buffer is recommended (see Note 9).
3.1.5 Flow Rate of the For the Model 491 Prep Cell the manufacturer recommends a flow
Collection Buffer and rate of 0.75–1.0 mL/min and a fraction volume of 2.5 mL for the
Fraction Volume Collected collection buffer. For the Mini Prep Cell a flow rate of 75–100 μL/
min with a fraction volume of 200–250 μL is recommended. We
recommend to use the lower end of these limits to reduce sample
dilution (see Note 10).
3.2 Preparative A typical application using preparative SDS-PAGE is presented, the

SDS-PAGE in Practice: purification of 25 kDa porcine amelogenin (P173) to single band
Purifying the “25 kDa” purity as determined by silver-stained analytical SDS-PAGE. In this
Porcine Amelogenin example we have used the Mini Prep Cell but our previous publica-
(P173) Using the Mini tion detailing the purification of recombinant amelogenin [17]
Prep Cell shows a real-world example of the larger Model 491 Prep Cell in
use.
1. For this demonstration, we took a lower deciduous porcine
incisor, at a stage just prior to completion of the crown, and
removed the enamel organ and the pulp.
2. The tooth was briefly rinsed in ultrapure water, blotted dry, and
crushed up in 300 μL of nonreducing SDS-PAGE sample
loading buffer using a glass rod.
3. After centrifugation to remove particulates, 15 μL of the sam-
ple was loaded onto the Mini Prep Cell.
4. A 12% gel was cast, 7 cm in height and topped with a 1 cm high

4% stacking gel. The ratio of acrylamide:bis-acrylamide in both
resolving and stacking gels was 37.5:1.
5. Electrophoresis was carried out according to the manufac-
turer’s recommendations at 1 Watt (constant power).
6. Once the bromophenol tracker dye had reached the bottom of
the gel fractions were collected at a flow rate of 80 μL/min and
200 μL fractions were collected.
7. Eighty fractions were collected and to identify which fractions
contained the 25 kDa amelogenin (P173) a 10 μL aliquot of
every third fraction was taken and freeze-dried.
8. The freeze-dried samples were dissolved in 10 μL SDS sample
loading buffer. A 1 μL aliquot of this was taken and diluted 10
times in SDS sample loading buffer and 5 μL of each fraction
subjected to analytical SDS-PAGE.
9. A small aliquot of the original 300 μL sample was diluted in
1:200 in SDS sample-loading buffer and 3 μL loaded.
10. The resulting gel was then silver stained (Fig. 4a).
11. From this gel we can establish that the target protein was
collected between fractions 55 and 64.
12. Each of the fractions in this range was then run as described
above (except that the samples were not diluted) to show the
protein contained in each fraction (Fig. 4b).
13. From this result we can see that there is a degree of overlap
between the 25 kDa amelogenin (P173) and 23 kDa amelo-
genin (P161) in fractions 55–58 so these contaminated frac-
tions would be excluded and fractions 59–64 pooled to obtain
the purified 25 kDa amelogenin (P173). Essentially, we have
sacrificed complete separation in fractions 55–58 for speed (see
Note 11).
14. The pooled fractions are comprised of the 25 kDa amelogenin
(P173) in collection buffer (SDS running buffer) and may need
to be desalted or buffer exchanged to recover the protein in a
form suitable for downstream applications (see Note 12).
15. The freeze-dried protein is then ready for downstream applica-
tions or a further round of purification using a second purifica-
tion methodology if a final polishing step is required.
16. Figure 3a also shows that this single run also achieved the
separation of the 23 kDa and 20 kDa amelogenins (see
Note 13).
Fig. 4 (a) Analytical SDS-PAGE of every third fraction collected during the purification of the target protein, the
25 kDa amelogenin (P173). The lanes labeled “Sample” show the starting material as loaded on the Mini Prep
Cell. Note that in addition to the target protein (boxed in red), the 23 kDa amelogenin (P162) and the 20 kDa
amelogenin (P148) have also been well resolved. In contrast the lower molecular weight proteins (exemplified
by the protein in fraction 25 boxed in blue) are less well resolved. This is because a higher percentage gel
would be required to resolve these smaller proteins. (b) Analytical SDS-PAGE of every fraction between
fractions 55 and 64 showing that the target protein is adequately resolved in fractions 59–64. These fractions
are pooled and desalted ready for use in downstream applications or additional round of purification if required
4 Notes
1. If the resolving gel and stacking gel are poured into a gel tube
that is not perfectly vertical the upper surface of the gel will not
be parallel with the bottom of the gel (the surface from which
the protein elutes).
2. Air bubbles can be avoided by degassing gel solutions and if
present should be dislodged by tapping the gel before
polymerization.
3. This is not a major issue with the Mini Prep Cell as the relatively
large surface area-to-volume ratio of the gel tube provides
sufficient cooling. However, when using the Model 491 Prep
Cell water should be pumped through the cooling core to
dissipate heat. The goal is to achieve a zero-temperature differ-
ential between the cooling core and the wall of the tube gel; we
do this by circulating water from a 3 L container that has been
at ambient temperature alongside for several hours.
4. We follow the manufacturer’s recommendations and allow the
resolving gel to polymerize overnight before pouring the
stacking gel.
5. Relative mobility ¼ distance migrated by the target protein
divided by the distance migrated by the ion front
(or bromophenol blue).
6. However, if the target protein differs in molecular weight from
its nearest contaminant by <10% it is advised to run a few mini
slab gels using a range of monomer concentrations to finalize
the concentration to be used.
7. Time on optimizing gel porosity is well spent as this is a most
critical parameter.
8. Although shorter gels mean a shorter run time, they may not
provide sufficient resolution as the target band may not be fully
separated from its closest contaminant at the point of elution
from the bottom of the gel. The obvious solution is to increase
the degree of separation by increasing gel length but overly
long gels and longer run times will increase band diffusion
during the run and band broadening. This could actually
reduce the degree of separation between the target protein
and its nearest contaminants.
9. The manufacturer’s instructions provide guidance on how to
eliminate these bubbles but essentially it is a case of gently
pushing collection buffer (using a large syringe) through the
tubing that normally carries collection buffer away from the
elution chamber. The whole assembly is tilted so that the
collection buffer feed line is uppermost so that bubbles float
upwards and are pushed out of the chamber through the feed
line. Tapping the assembly against the bench helps dislodge any
stubborn bubbles.
10. We have not encountered any difficulties using the lower flow
rates but one can envisage that very low flow rates may increase
the risk of proteins electrophoresing through the dialysis mem-
brane or being mixed with close contaminants that begin to
elute before the target protein has been fully cleared from the
elution chamber.
11. Increasing the gel height from a relatively short 7 cm to 10 cm

would be an obvious step to achieve a better separation in
fractions 55–58 but this would increase the run time.
12. A number of options are available to recover the protein, e.g.,
simple dialysis, centrifugal ultrafiltration, or gel filtration chro-
matography. We favor gel filtration chromatography using
appropriately sized Sephadex G25 columns using 0.125 M
formic acid as the mobile phase. The formic acid maintains
amelogenin solubility during desalting, and being volatile is
easily removed by freeze-drying.
13. To achieve the simultaneous purification of these three amelo-
genins in a single step using other purification techniques
would be challenging (at least in our hands) and this demon-
strates the potential efficiency gains that are possible using
preparative electrophoresis. It is also apparent from Fig. 3a
what happens if an inappropriate gel porosity is used. If one
of the lower molecular weight proteins present in fraction
25 was the target in this exercise, then the 12% gel used here
failed to achieve separation. To separate these lower molecular
weight proteins a higher percentage gel would have been
required.
5 Conclusion
Preparative SDS-PAGE may be a useful addition to the range of

techniques used to purify developing enamel proteins (and proteins
from other tooth tissues). We are currently exploring the combina-
tion of preparative SDS-PAGE as described here with preparative
native charge PAGE. In effect, this would provide a 2D separation
of the target protein based on molecular size and the native charge
carried by the protein. Such an approach would be ideal for separ-
ating for example non-phosphorylated 25 kDa amelogenin (P173)
generated in the matrix by the action of phosphatase enzyme from
the nascent form of the protein that is phosphorylated at serine 16.
Phosphorylation would not affect the migration of the 25 kDa
amelogenin (P173) during SDS-PAGE, but would affect its migra-
tion during native PAGE since at the appropriate pH the phos-
phorylated isoform would be carrying an additional negative charge
due to the presence of the –PO43 group.
Acknowledgments
Claire Gabe was supported by a University of Leeds 110 Anniver-

sary Research Scholarship.
References
1. Mechanic GL, Katz EP, Glimcher MJ (1967) In: Kobayashi I, Ozawa H (eds) Biomineraliza-
The Sephadex gel filtration characteristics of tion: formation, diversity, evolution and appli-
the neutral soluble proteins of embryonic cation; Proceedings of the 8th International
bovine enamel. Biochim Biophys Acta Symposium on Biomineralization; Niigata,
133:97–113 Japan; Hadano, Japan: Tokai University Press;
2. Belcourt AB, Fincham AG, Termine JD (1983) 2004
Bovine high molecular weight amelogenin pro- 11. Yamakoshi Y, Tanabe T, Fukae M, Shimizu M
teins. Calcif Tissue Int 35:111–114 (1994) Porcine amelogenins. Calcif Tissue Int
3. Eggert FM, Allen GA, Burgess RC (1973) 54:69–75
Amelogenins. Purification and partial charac- 12. Simmer JP, Fukae M, Tanabe T, Yamakoshi Y,
terization of proteins from developing bovine Uchida T, Xue J, Margolis HC, Shimizu M,
dental enamel. Biochem J 131:471–484 Dehart BC, Hu CC, Bartlett JD (1998) Purifi-
4. Fincham AG (1979b) A simple procedure for cation, characterization, and cloning of enamel
the isolation of a major amelogenin polypep- matrix serine proteinase 1. J Dent Res
tide component. Biochem J 181:171–175 77:377–386
5. Fukae M, Ijiri H, Tanabe T, Shimizu M (1979) 13. Yamakoshi Y, Hu JC, Zhang H, Iwata T, Yama-
Partial amino acid sequences of two proteins in koshi F, and Simmer JP (2006) Proteomic
developing porcine enamel. J Dent Res analysis of enamel matrix using a two-dimen-
58:1000–1001 sional protein fractionation system. Eur J Oral
6. Shimazu S, Fukae M (1983) Enamel proteins. Sci 114 Suppl 1, 266–271. https://doi.org/
In: Suga S (ed) Mechanisms of tooth enamel 10.1111/j.1600-0722.2006.00279.x
formation. Quintessence Publishing Co, Inc, 14. Fincham AG (1979a) The amelogenin prob-
Tokyo lem: a comparison of purified enamel matrix
7. Hu CC, Fukae M, Uchida T, Qian Q, Zhang proteins. Calcif Tissue Int 27:65–73
CH, Ryu OH, Tanabe T, Yamakoshi Y, 15. Fukae M, Tanabe T (1987) Nonamelogenin
Murakami C, Dohi N, Shimizu M, Simmer JP components of porcine enamel in the protein
(1997) Sheathlin: cloning, cDNA/polypeptide fraction free from the enamel crystals. Calcif
sequences, and immunolocalization of porcine Tissue Int 40:286–293
enamel sheath proteins. J Dent Res 16. Brookes SJ, Kirkham J, Lyngstadaas SP, Shore
76:648–657 RC, Wood SR, Robinson C (2000) Spatially
8. Hu JC, Yamakoshi Y, Yamakoshi F, Krebsbach related amelogenin interactions in developing
PH, Simmer JP (2005) Proteomics and genet- rat enamel as revealed by molecular cross-
ics of dental enamel. Cells Tissues Organs linking studies. Arch Oral Biol 45:937–943
181:219–231 17. Gabe CM, Brookes SJ, Kirkham J (2017) Pre-
9. Limeback H, Simic A (1990) Biochemical parative SDS PAGE as an alternative to
characterization of stable high molecular- His-Tag purification of recombinant amelo-
weight aggregates of amelogenins formed dur- genin. Front Physiol 8:424
ing porcine enamel development. Arch Oral 18. Sutton RMC, Stalcup AM (2000) Preparative
Biol 35:459–468 electrophoresis. In: Wilson ID, Adlard ER,
10. Yamakoshi Y, Hu JCC, Ryu OH, Tanabe T, Cooke M, Poole CF (eds) Encyclopedia of sep-
Oida S, Fukae M, et al. (2001) A comprehen- aration science. Academic Press, San Diego,
sive strategy for purifying pig enamel proteins. London
Chapter 26
Using ImageJ (Fiji) to Analyze and Present X-Ray CT

Images of Enamel
Steven J. Brookes
Abstract
X-ray micro CT has become a popular methodology for the nondestructive analysis of dental tissues and has
been used extensively in the amelogenesis field. The aim of this chapter is to introduce ImageJ/Fiji to
researchers new to CT scanning and the analysis of CT image data. The program can be applied to analyzing
X-ray CT images of enamel but can be extrapolated to other tissues as well.
Key words X-ray micro CT, Dental tissues, Enamel, ImageJ/FIJI
1 Introduction
X-ray micro CT has become a popular methodology for the non-

destructive analysis of dental tissues and has been used to generate
data as reported in several publications in the amelogenesis field
[1–6]. Given the relevant software tools, it is relatively straightfor-
ward to extract data relating to enamel mineral density, volume, and
thickness. A common scenario for researchers using X-ray micro CT
for the first time might be that their institute’s CT scanning facility
presents them with a portable hard disk carrying hundreds or
maybe thousands of CT images that represent a digitally sectioned
model of their specimen. In some cases, the necessary software and
support will be provided for the researcher to manipulate and
analyze their CT images in the digital domain and extract the
required data. On the other hand, support might be limited, and
the researcher might be left with several gigabytes of image data but
have no resources available to analyze the images. Fortunately,
there are free alternatives to expensive software used to analyze
CT images, one of which, ImageJ [7], is the subject of this chapter.
Electronic supplementary material: The online version of this chapter (https://doi.org/10.1007/978-1-4939-

9012-2_26) contains supplementary material, which is available to authorized users.
267
268 Steven J. Brookes
The aim of the chapter is to introduce ImageJ to researchers new to

CT scanning and the analysis of CT image data. A huge amount of
material relating to ImageJ is available online, and together with an
active user forum group (http://forum.imagej.net/), these
resources are very useful for solving many image analysis problems.
With the increasing availability of high-resolution X-ray CT scan-
ning facilities in institutions, we all need to become more familiar
with image analysis, and it is hoped that this chapter will help
researchers new to analyzing CT images to extract useful data
from their CT images.
2 Materials
2.1 Software: FIJI Throughout the chapter, the so-called Fiji derivative of ImageJ is
Derivative of ImageJ used as this version comes preloaded with many third-party plugins
[8], some of which are demonstrated here. Fiji can be freely down-
loaded as a zipped folder from https://imagej.net/Fiji/
Downloads#Installation. Fiji is not installed as such. It is down-
loaded as a portable application in a zipped folder. Once unzipped,
the Fiji.exe file can be run directly from the executable file. As with
all digital image-handling operations, the computer hardware
needs to be up to the task especially with regard to the amount of
RAM available which should be greater than the size of the image
stack under analysis for best performance. The screenshots pre-
sented here were obtained by running Fiji on a Microsoft Windows
PC, though Fiji runs in a similar manner on Apple-based systems.
Instructions are provided in the text that relate to selecting various
menu choices. For example, the instruction to open a file would be
“File>Open.” In other words, click on the “File” menu and then
the submenu item “Open.” Instructions relating to using one of
the tools on the tool bar simply describe the tool in question, e.g.,
“Rectangular selection” tool. There is a wealth of online documen-
tation for ImageJ (which applies to Fiji), and the ImageJ User
Guide at https://imagej.nih.gov/ij/docs/index.html is particu-
larly useful as it describes the user interface, various tools, and
menu commands.
2.2 Getting Started 1. On first opening of FIJI, the user is presented with a decep-
tively simple looking interface that belies the flexibility of the
program.
2. To open a reconstructed CT image stack, use “File>Import>-
Image Sequence.”
3. The user clicks on one of the images comprising the stack
which opens the import dialogue box where the range of
images to be imported can be selected as well as several other
options whose function can be seen by clicking the Help but-
ton, which links to relevant online help files (see Note 1).
Image J/FIJI & X-Ray CT Scans 269
4. Once the image stack is open, the user can scroll through the
CT slices using the slider at the bottom of the window.
3 Methods
3.1 Basic Operations 1. Simple operations can be carried out on the stack using a range
on Image Stacks of options under “Images>Stacks.” For example, if a line is
drawn on the image using the “Straight line selection” tool
from the top of the window downward, then the Reslice com-
mand accessed through “Images>Stacks>Reslice” can be used
to produce a new image stack orthogonal to the original stack
(see Fig. 1) (see Note 2).
2. Selecting “Images>Stacks>Orthogonal Views” opens two
additional windows that display YZ and XZ views in addition
to the original XY view. Clicking in any of the three planes
updates the other two windows such that the cross hairs always
intersect at the same point in the sample.
In the example shown in Fig. 1, a rodent mandible has been

resectioned longitudinally (sagittal section) to give a result
Fig. 1 Using the “Reslice” command on an image stack to generate an

orthogonal stack. A straight-line section drawn on the stack defines the new
plane of section. The number of new sections generated is determined by the
“Slice count” entered by the user in the “Reslice” dialogue box
Fig. 2 Using “Reslice” and “Z-Projection” to achieve a section along the full length
of the curved rodent mandibular incisor. (a) The stack is resliced using (Images>-
Stacks>Reslice); in this case reslicing can be carried out from the top to bottom to
generate a new stack looking down on the molars. (b) The new stack generated by
reslicing operation carried out in a. (c) A Z-projection of stack (b) is generated
(Images>Stacks>Z-Projection) which reveals the curved path of the incisor. The
“Segmented” or “Freehand” line tool is used to draw a selection line following
the path of the incisor. (d) The line selection drawn in (c) is transferred to stack
(b) by selecting the stack and using the “Restore Selection” command
(Edit>Selection>Restore Selection). (e) Finally, Images>Stacks>Reslice is
used to generate a section along the whole length of the incisor
3.2 How to Resection analogous to that obtained if the mandible had been sectioned
Longitudinally a longitudinally using a conventional microtome. As can be seen
Rodent Mandible from the example, it is not possible to longitudinally section the
Image Using FIJI (See mandible so that the full length of the incisor is visible. This is
Fig. 1) because the rodent incisor is curved and runs in and out of the
section obtained. However, using Fiji it is possible to reslice along a
curve in order to obtain sections that would be impossible to obtain
using a conventional microtome.
In order to generate a section that shows the full length of the
rodent incisor using Fiji, as illustrated in Fig. 2 (see Fig. 2), follow
the following steps:
1. The first step is to reslice the stack from the top so a view is
obtained looking down onto the molars. This is easily achieved
by using the “Images>Stacks>Reslice” command and opt to
reslice the stack from the bottom (or top—it makes little dif-
ference this case). On clicking OK, the user can see the progress
of the operation as each reslice is indicated by a yellow line that
is drawn across the rectangle. Once the reslicing operation is
complete, a new stack is generated.
2. In order to see the curved path of the incisor, the user generates
a Z-projection of this new stack which is essentially a composite
image of all the slices in the stack. The Z-projection is gener-
ated using “Images>Stacks>Z Project.” Several projection
types are available, but the “Sum slice” option works well.
The curved path of the incisor is clearly visible on the
Z-projection image, and the “Free hand” tool or “Segmented
line” tool (accessed by right clicking on the “Straight line
selection” tool in the bottom right-hand corner) can be used
to draw a line that follows the curve of the incisor and defines
where the section will be cut. The line is treated as a selection
by Fiji and can be placed on the original stack used to generate
the Z-projection image. This is simply a matter of selecting the
stack of images and applying the selection to the stack using the
“Edit>Selection>Restore Selection” command. This transfers
the line drawn on the Z-projection onto the image stack.
3. Next, the “Images>Stacks>Reslice” command is used to
reslice the stack along the line of the curved selection which
generates a longitudinal image of the whole incisor that would
be impossible to obtain using a conventional microtome.
4. Although the ability to reslice CT image stacks is very useful for
understanding the structure and internal architecture of a spec-
imen, Fiji provides all the tools needed to carry out quantitative
assessment of mineral density as the grayscale value of the
image is proportional to its X-ray attenuation coefficient (see
Note 3). The first task when assessing mineral density of
enamel is to isolate, or segment, the enamel from other tissues
such as the dentine and bone. Using image analysis parlance,
the enamel needs to be identified as a region of interest (ROI).
3.3 Selecting a CT slices of teeth may include enamel, dentine, bone, and soft
Region of Interest tissues. To measure the density of the enamel or the area of enamel
(ROI) for Quantification present in an image slice, it is necessary to select a region of interest
Purposes (ROI) that delineates just the enamel (see Note 4). Once an ROI is
defined, the mean grayscale value of the pixels bounded by the ROI,
the area of the ROI, and a number of other parameters associated
with the ROI can then be recorded in a “Results” window using the
“Analyze>Measure” command (or keyboard shortcut M).
An example is shown in Fig. 3a where an ROI has been hand
drawn around the enamel on a transverse slice through a rodent
Fig. 3 Defining a region of interest (ROI) using manual selection by hand and manual thresholding. (a) The
“Polygon” selection tool has been used to hand draw an ROI around the enamel present on a rodent incisor.
Once the ROI is defined, a number of parameters associated with the pixels within the ROI can be measured
using the “Analyze>Measure” command (or keyboard shortcut M). The data are added to a “Results” window.
Here the area of the ROI and the mean grayscale value ( standard deviation) of the pixels it contains have
been recorded, but numerous other measurements can be set using the “Analyze>Set Measurements”
command. (b) As an alternative to manually drawing an ROI, the enamel can be segmented using the
“Image>Adjust>Threshold” dialogue box (inset). Here, setting the threshold range to between 125 and
250 highlights pixels comprising the enamel but also highlights some regions (speckling) in the dentine. (c)
The thresholded enamel can be selected using the “Wand Tool.” (d) Once selected with the “Wand Tool,” the
enamel can be measured using the “Analyze>Measure” command (or keyboard shortcut M)
incisor using the “Polygon selection” tool. This hand-drawn ROI

has an area of 1686 pixels and a mean grayscale value of
157.6 17.7 as recorded in the “Results” window (Fig. 3a inset)
(see Note 5). If ROIs are defined manually, then it is certainly a
good practice to statistically determine interoperator precision and
intra-operator coefficient of variation as previously described
[6]. An alternative to manually drawing ROIs is to use the thresh-
olding function in Fiji.
3.4 Image At its simplest, thresholding selects all those pixels in an image that
Thresholding falls within a certain predefined grayscale range. This is easily
achieved in Fiji using “Image>Adjust>Threshold.” This opens
the “Threshold” dialogue box in which the sliders can be used to
set the upper and lower grayscale limits that will dictate which
image pixels are thresholded. In Fig. 3b the upper and lower
thresholds have been manually set to 255 and 125, respectfully,
and those pixels having a grayscale value within this range have been
colored red (this is an 8-bit image with 256 levels of gray (0–255)
(see Note 6).
It is possible to measure the thresholded regions directly by
selecting the “Limit to threshold” tick box in the “Set measure-
ments” window, which is accessed using “Analyze>Set measure-
ments.” However, the data obtained would include contributions
from any thresholded pixels present in the dentine and mandibular
bone. This is clearly not desirable; instead it is better to generate an
ROI that is defined by the thresholded area of interest—the
enamel. The ROI is easily generated by clicking in the thresholded
enamel with the “Wand (tracing)” tool as shown in Fig. 3c. Hold-
ing the shift key down while using the “Wand (tracing)” tool allows
multiple thresholded areas to be selected which is useful if there are
separate areas of enamel present on the CT slice. Once the thre-
sholded enamel has been delineated by an ROI, it is a good idea to
remove the thresholding using the “Reset” button in the thresh-
olding dialogue box (Fig. 3d) (see Note 7). If the ROI is satisfac-
tory, measurements are recorded using “Analyze>Measure”
(or keyboard shortcut M). As can be seen from the Results window
in Fig. 3d (inset), the data obtained is, not surprisingly, slightly
different to the data shown in Fig. 3a which was obtained using a
hand-drawn ROI (see Note 8).
3.5 Auto Auto thresholding removes operator bias in selecting a ROI. Auto-
Thresholding with Fiji thresholding techniques apply algorithms that do far more than
simply select a pixel based on its grayscale value. Fiji allows users to
try 16 different auto-thresholding algorithms in an attempt to
segment an image (see Note 9). Before attempting to perform
auto thresholding, we need to decide which of the 16 thresholding
algorithms available in Fiji is best suited to the application in
question. There are no rules here, and it is a case of trying each
Fig. 4 Defining a region of interest (ROI) using auto thresholding. (a) A CT slice through a rodent molar ready for
auto thresholding using “Image>Adjust>Auto Threshold.” (b) Auto thresholding trialing all 16 thresholding
algorithms generates a binarized collage showing the thresholding obtained by each algorithm. Only four
algorithms were able to threshold the enamel with any degree of success, and even in these cases, speckling
was present in the dentine. (c) Thresholding obtained by rerunning “Image>Adjust>Auto Threshold” with the
“Yen” algorithm selected. (d) The required ROI is generated using the “Wand Tool” (see text for details) prior to
making the measurement
algorithm to see which gives the best segmentation of the enamel.

Fortunately, Fiji provides a simple way to quickly identify the best
algorithm for the job. A CT slice through a mouse molar is shown
in Fig. 4a.
After selecting the image, “Image>Adjust>Auto Threshold” is
used to open the Auto Threshold dialogue window. The default
option is to try all the algorithms, and selecting OK will generate a
collage comprising 16 binary images showing the results obtained
using each of the 16 auto-threshold algorithms (Fig. 4b). From this
initial screen, it is clear that only four of the algorithms (MaxEn-
tropy, RenyiEntropy, Shanbhag, and Yen) managed to segment the
enamel, but even they erroneously segmented some pixels in the
dentine generating a “speckling” effect. To apply 1 of the 16 algo-
rithms to the image, click on the image, and run “Ima-
ge>Adjust>Auto Threshold” again. This time a specific
algorithm is selected from the drop-down menu (Yen in this
case), and the option “SetThreshold instead of Threshold (single
images)” is selected. On clicking OK, the thresholded pixels will be
highlighted red on the image in (Fig. 4c). The ROI of interest can
then be generated using the “Wand (tracing)” tool. However,
selecting the thresholded area with the Wand is a little more
involved than the previous example as the enamel surrounds the
dentine which represents an area that is not part of the ROI. In this
case, select the Wand (tracing) tool, and click in the thresholded
enamel. This draws an ROI around the whole tooth. To deselect
the dentine, press the ALT key down, and click the Wand anywhere
in the unthresholded dentine area. This should leave just the
enamel selected that can then be measured using “Analy-
ze>Measure” (or keyboard shortcut M).
3.6 Using Macros to The method described above using the “Wand (tracing) tool” is
Automate fine for generating ROIs by hand if there are only a couple images
Thresholding and to deal with. However, using this method to threshold and select
Generating ROIs on ROIs on multiple slices would be extremely labor intensive. In such
Single Images cases it would be far more efficient to use Fiji’s macro language to
automate the process. In cases similar to the one illustrated above,
where thresholding is not 100% accurate resulting in speckling, the
“Analyze Particles” plugin can be used to despeckle the image
leaving just the enamel thresholded which can then be selected
automatically by Fiji to generate the required ROI. As shown in
Fig. 5, several steps are required to carry out despeckling using the
“Analyze Particles” plugin by hand, but these can all be executed by
running an ImageJ macro script.
The macro illustrated below was compiled using the “Macro
Recorder” (“Plugins>Macros>Record”) with a little additional
code added to the script to provide additional functionality (see
Note 10).
In the example shown here, the macro does the following:
1. Gets the name of the newly opened image.
2. It then runs “Auto Threshold” and sets the threshold (in this
case using the Yen algorithm).
3. It then runs “Analyze Particles” and selects all objects bigger
than 500 pixels and generates a binary image (mask) of those
objects. Depending on the image in question, this value can be
changed to optimize the macro’s function so that only the
enamel is recognized.
4. It then runs “Create Selection” which selects the enamel in the
binarized image.
5. It then closes the window showing the binarized image.
6. It makes the original image window the focus of the program.
7. The selection created in step 4 is then applied (restored) to the
original image to generate the ROI.
8. The threshold is reset, so it is easy to see exactly what has been
selected.
9. The measurements (area, mean grayscale, and standard devia-
tion in this case) are recorded to three decimal places in a
“Results” window from where they can copied to Excel, etc.
Fig. 5 Using macros to automate thresholding and generate ROIs. The “Wand Tool” is fine for generating ROIs,
but the method is labor intensive, and analyzing multiple slices would be time consuming. Here the “Analyze
Particles” tool is used to eliminate speckling which means no human intervention is required to identify and
manually select the enamel in isolation from any speckling. This means a macro can be used to auto threshold
the enamel, despeckle the image, generate the ROI, select the only thresholded object present (the enamel),
and carry out the measurement. The macro (see text for code and detailed mode of operation) carries out
steps a–h shown in the figure without user input. (a) A CT slice through a rodent molar. (b) The macro runs
auto thresholding using the “Yen” algorithm (the macro can be edited to run other algorithms to suit the image
in question). (c) The resulting thresholded image exhibits speckling in the dentine. (d) The “Analyze Particles”
tool generates a binary image (e) comprising only particles (objects) greater than 500 pixels in size—i.e., the
enamel (note, the particle size is set by the macro, and these values can be edited to suit the image in
question). (f) The macro runs “Create selection” which selects any objects present. (g) The macro reselects
the image window (a) and runs “Restore selection” which generates the ROI by copying the selection
generated in (f) onto the CT slice. (h) The macro sets which measurements are to be made (area and mean
grayscale standard deviation) and the number of decimal places to be used, makes the measurements, and
presents them in a “Results” table
To run the macro, it is simply a matter of saving the following

script as a .txt file and then using “Plugins>Macros>Run” to
open and run the macro. Alternatively, the macro can be installed
using Plugins>Macros>Install and will then appear in the
Plugins>Macros drop down menu (see online video for example).
//Macro 1
name=getTitle;
run("Auto Threshold", "method=Yen white setthreshold");
run("Analyze Particles. . .", "size=500-Infinity show=Masks");
run("Create Selection");
run("Close");
selectWindow(name);
run("Restore Selection");
resetThreshold();
run("Set Measurements. . .", "area mean standard redirect=None
decimal=3");
run("Measure");
The actions listed above are then executed immediately in a

fraction of a second (see Note 11).
3.7 Segmentation Fiji includes the “Trainable Weka Segmentation” tool [9] which
Using a Machine uses the image analysis tools in Fiji to feed data into the Waikato
Learning Approach Environment for Knowledge Analysis (Weka) data mining software
in Fiji platform [10]. The machine learning algorithms and data prepro-
cessing tools available in Weka enable users to train the Weka
segmentation tool by providing it with examples that allows a
pixel to be classified in terms of whether or not it belongs to a
specific population or class of pixels. The user “trains” the tool by
manually delineating the different structures (e.g., enamel, dentine,
and background) by manually drawing ROIs on each structure.
ROIs placed on enamel would contain pixels of one class, ROIs
placed on dentine would contain pixels of a second class, and so
on. Each of these user-defined classes is then interrogated using the
numerous Fiji image analysis algorithms and filters. In essence, the
plugin mines the data generated and identifies specific characteris-
tics that can be used to distinguish between the different classes of
pixel. Depending on the image size, the image complexity, and the
number of Fiji image processing routines and filters assigned to the
training task, training may take some time even when using a high-
specification PC. However, once the plugin has successfully
“learned” how to distinguish between the different areas or classes
identified by the user, the resulting “Classifier model” (in effect the
lesson learned) can be saved and applied directly to similar images
which can then be segmented immediately with no further training
required. Figure 6 illustrates how the Trainable Weka Segmentation
tool is used to segment enamel.
In this case, the tooth is from a patient carrying a mutation in
the amelotin gene, and the enamel is undermineralized compared
to control enamel which makes segmenting the enamel from the
dentine more challenging [11]. The first step is to open the image
to be analyzed (in this case slice 285 of an image stack). The
Trainable Weka Segmentation tool is opened through “Plu-
gins>Segmentation>Trainable Weka Segmentation,” and the
slice is automatically loaded into the plugin window. The user
then uses one of the drawing tools to draw around populations of
pixels that exemplify what our eyes and experience allow us to
Fig. 6 Using a machine learning approach to thresholding and generating ROIs. The “Trainable Weka
Segmentation” tool employs a machine learning approach to segmentation which may greatly improve
thresholding efficiency, especially if the enamel is poorly mineralized and contains areas exhibiting a
grayscale value similar to that of dentine. This figure shows the basic methodology involved when using
the “Trainable Weka Segmentation” tool. (a) Step 1: A CT slice (slice 285) through a human molar exhibiting
areas or poorly mineralized enamel is opened in Fiji and is automatically loaded into the “Trainable Weka
identify as the enamel, dentine, and background (enamel shaded

red, dentine shaded green, and background shaded purple) (see
Note 12). The three areas of pixels exemplifying enamel (red) are
added to class 1, the single area of pixels exemplifying dentine
(green) are added to class 2, and the three areas of pixels exemplify-
ing the background (purple) are added to class 3. The user then
clicks “Train classifier,” and the program then “learns” by identify-
ing features that are able to differentiate between each of the three
classes of pixels (enamel, dentine, and background) that the user
identified. Once the training is complete, the user clicks “Create
result,” and a new window called “Classified image” is generated
which shows all the pixels comprising the image according to
whether they have been classified as enamel (red), dentine
(green), or background (purple). There is no speckling in the
dentine, and the enamel (red) has been successfully segmented.
An ROI can then be generated around the red enamel using the
“Wand (tracing)” tool as described above, and the resulting ROI
can then be overlaid on the original slice by clicking on the window
showing the original slice followed by “Edit>Selection>Restore
Selection.” The enamel can then be measured using “Analyze
Measure” (or keyboard shortcut M). If the “Classifier model”
generated successfully segments the target tissue, it can be saved
(using “Save classifier”) ready to apply to a new slice.
Figure 6b shows the result of applying the “Classifier model”
generated using slice 285 to another slice in the same stack (slice
308 in this example). It is simply a case of opening the new slice and
loading it into the Trainable Weka Segmentation tool by running
the plugin as described above. No training is required; instead the
“Classifier model” saved above is loaded (using “Load classifier”),
and a new classified image window is generated by clicking “Create
Fig. 6 (continued) Segmentation” tool on opening the tool. The user then draws around areas representative
of the enamel (red), dentine (green), and background (purple). It is important to note that the user is not
attempting to draw an ROI as such but is simply providing the tool with examples of the three different pixel
classes present (these classes being pixels belonging to enamel, dentine, and background). As each area is
delineated, it is added to the corresponding class using the “Add to class” buttons on the right-hand side of the
window. Step 2: The user clicks the “Train classifier” button, and the tool attempts to find features that can be
used to differentiate the three different classes of pixels and generate a classifier model. Training may take
some time, but on completion the resulting segmentation is overlaid on the image (not shown), and if
acceptable the user clicks “Create result,” and a new window opens showing the so-called Classified image.
As shown here, the Wand Tool has been used to select the red area (enamel). Step 3: the selection has been
copied to the original image ready for the measurements to be carried out. (b) It is not always necessary to
train the tool for every new image analyzed. Here, a different slice (slice 308) from the same tooth has been
analyzed using the classifier model generated during training using slice 285. Reusing a pre-existing classifier
model in this way speeds up analysis as it negates having to carry out training—a process that can take some
time depending on the image in question and the range of image analysis filters and algorithms the tool uses
(selected under “Settings”) in an attempt to differentiate between the pixel classes defined by the user
result.” A region of interest is created and overlaid on the original

slice ready for measurements to be recorded.
3.8 How Does the The auto-thresholding algorithms are simple to use and make
Trainable Weka relatively little demand on computing power. In contrast, mastering
Segmentation Tool’s the “Trainable Weka Segmentation” tool takes some effort and
Implantation of depending on the image size and the battery of Fiji image analysis
Machine Learning tools selected to classify the pixels requires more computer run
Compare to Using the time. However, the results obtained using trainable segmentation
Standard Auto- can be superior, especially where the enamel is not well mineralized
Thresholding and its grayscale value overlaps with that of the dentine or other
Algorithms tissues.
A comparison between auto thresholding (using the Yen algo-
rithm—the most efficient of the 16 algorithms in this case) and the
“Trainable Weka Segmentation” tool is shown in Fig. 7. Here,
another slice from the stack featured in Fig. 6 has been segmented
using both methods. The white boxes (i–iii) overlaid on the scans
highlight areas of enamel exhibiting grayscale values similar to the
dentine which can be difficult to segment manually or using the
standard auto-thresholding algorithms (Fig. 7a). Auto threshold-
ing did threshold the enamel, but the generated speckling in the
dentine and enamel and the magnified images of the boxed areas
show that auto thresholding failed to segment poorly mineralized
enamel due to its lower grayscale value (Fig. 7b). In contrast, the
“Trainable Weka Segmentation” tool handled the problem areas
more successfully (Fig. 7c). In simple terms, the tool has learnt
from the examples provided by the user that enamel can contain
poorly mineralized areas and can then differentiate these areas from
dentine. The dentine itself is not speckled because again the tool
has learnt that dentine can include small areas of pixels having a
relatively high grayscale value (see Note 13).
3.9 Quantification of So far we have largely concentrated on means of segmenting the

CT Images: Estimating enamel so that the mean grayscale value for enamel can be measured
Mineral Density and in isolation from the other tissues. We now consider how Fiji can
Adding Scale Bars utilize successfully segmented enamel to estimate enamel mineral
density by calibrating the grayscale images against suitably propor-
tioned hydroxyapatite standards of known mineral density. It is
essential that calibration scans are acquired under the same scan-
ning and image reconstruction parameters. The simplest way to
ensure this is to scan specimens and standards together. We gener-
ally use a three-point calibration comprising cylindrical hydroxyap-
atite standards with densities of 0.25, 0.75 (Bruker, Kontich,
Belgium), and 2.9 g/cm3 (Himed, Bethpage, NY, USA) as
described by Schmitz et al. [6]. Mature enamel has a density that
can exceed 3 g/cm3, and we acknowledge that we may have to
extrapolate to ascribe an estimated mineral density to the most
highly mineralized enamel.
Fig. 7 Comparing segmentation achieved using standard auto thresholding (using the Yen algorithm) and
“Trainable Weka Segmentation” tool. (a) A CT slice exhibiting hypomineralized enamel with several problem-
atic features indicated by boxed areas (i–iii) which are shown magnified to the right of the main image. (b)
Auto thresholding using the Yen algorithm (the most effective of the 16 algorithms available) failed to segment
Calibrating CT images for mineral density involves determin-

ing the mean grayscale value for each standard and plotting the
grayscale value against density to generate a calibration curve from
which unknown grayscale values can be converted to mineral den-
sity. Fiji provides a calibration routine that simplifies this process.
1. The first step is to measure the mean grayscale values of the
standards. Figure 8a shows an image stack window showing a
transverse slice through the 0.75 g/cm3 hydroxyapatite
standard.
2. An ROI is created by drawing free hand with the “Oval selec-
tions” to draw a circle.
3. Once the ROI is created, the mean grayscale value of the pixels
within the ROI is recorded using “Analyze>Measure” (or the
keyboard shortcut “M”), and the measurements are recorded
in a “Results” window (not shown).
4. The slider at the bottom of the window can be used to show
another slice, and a new ROI is created to record duplicate
mean grayscale values.
5. This process is repeated until multiple grayscale measurements
have been recorded for each standard (typically we measure the
mean grayscale value of five randomly selected slices for each
standard).
6. To relate these mean grayscale values to actual mineral density,
use the “Analyze>Calibrate” function. A new window opens
that lists the mean grayscale values recorded in the left-hand
column and an empty right-hand column in which the user
enters the corresponding known mineral densities of the stan-
dards—0.25, 0.75, and 2.9 g/cm3 (Fig. 8b).
7. Fiji then uses this data to construct a regression curve. A
number of curve-fitting options are available in the “Function”
drop-down box, but the “Straight Line” function fits the cali-
bration data we obtain. Ticking the “Global calibration box”
will apply the calibration to all images opened until Fiji is closed
down. The “Save” function allows the calibration data to be
saved.
Fig. 7 (continued) the poorly mineralized areas of enamel (i), (ii), and (iii). Note, in addition to speckling in the
dentine, there was speckling in the enamel where the density of the enamel was similar to that of the dentine.
(c) Segmentation using trainable segmentation achieved a much better result. Even the poorly mineralized
enamel was correctly segmented, and there was no speckling in either the dentine or the enamel. By
analyzing the different classes of pixels defined by the user, the tool “learnt” enamel may contain patches of
pixels with lowered grayscale value and conversely that dentine may contain pixels with higher grayscale
values and managed to generate a classifier model that was able to distinguish the enamel, dentine, and
background
Fig. 8 Calibrating grayscale values and image pixel size in terms of mineral density (g/cm3) and standard units
of length (μm). (a) To convert grayscale values to units of density, hydroxyapatite standards of known density
are scanned, and their mean grayscale is determined by manually drawing an ROI and recording the data in a
Results table (not shown) using “Analyze>Measure” (or keyboard shortcut M). (b) Once data for several
different standards has been acquired, “Analyze>Calibrate” is used to open the calibration dialogue box. The
measured data automatically appears in the left-hand column, and the user enters the corresponding known
mineral densities in the right-hand column. The user selects a curve-fitting function (straight line in this case).
The user can save these values so image(s) can be reanalyzed later without having to repeat the calibration
measurements. The calibration will be applied to all images opened during the session if the “Global
calibration” box is ticked. The calibration is applied by clicking OK. (c) If “Show plot” is ticked, a calibration
graph is also generated. This enables the user to assess the linearity of the calibration obtained. (d) Once the
calibration is applied, any grayscale measurements obtained will now be given in terms of mean mineral
density. (e) The Results table associated with the image shown in (d) shows the mean mineral density in g/cm3
with standard deviation rather than a grayscale value. (f) The set scale dialogue box opens and allows users to
calibrate images in standard units of length rather than pixels. The known image pixel resolution is entered
(6.32 μm in this case). Scale bars, as shown in (d), can then be added to an image using “Analy-
ze>Tools>Scale Bar” dialogue box (not shown)
8. The “Open” function can then be used to repeat any analysis in

the future without having to remeasure the standards.
9. Finally ensure that the “Show plot” box is ticked, and click
“OK” to show the resulting calibration plot (Fig. 8c). The plot
window gives an immediate visual impression of the curve
fitting and provides the coefficient of determination (R2),
0.9997 in this case, and the equation for the plot. All individual
data points are plotted in red, and the moveable cross hairs
allow the user to obtain the calculated Y value (mineral density)
for any X value (mean grayscale value).
10. The “List” function opens a new window (not shown) that
shows the calculated mineral density associated with every
grayscale value between 0 and 255 (for 8-bit images). To
measure enamel density, it is simply a matter of creating an
ROI over the enamel and recording the mean grayscale value
using “Analyze>Measure” (or the keyboard shortcut “M”).
The measurement will be recorded in the results window but
will appear as mean mineral density rather than just the mean
grayscale value.
11. Selecting “Results>Set measurements” in the Results window
will allow several other parameters to be included in the results
table (e.g., the standard deviation and the perimeter of the ROI
to name but two). A typical result using the calibration data
illustrated here is shown in Fig. 8d where the incisor enamel
and molar enamel present in a slice through a mouse mandible
have been thresholded, an ROI has been generated, and the
mean mineral densities (+/ standard deviation) have been
determined and shown in the Results table (Fig. 8e) (see
Note 14).
The other calibration that can be carried out is to set the scale
by converting pixels to conventional units of length. This allows
users to generate an appropriately calibrated scale bar as seen in
Fig. 8d.
1. To set the scale, select “Analyze>Set Scale” to open the “Set
Scale” dialogue window (Fig. 8f).
2. The distance in pixels is set to 1, and the size of a single pixel
(image resolution usually available in a log file generated during
scanning or image reconstruction) is entered under known
distance. In the example shown, 1 pixel represents 6.32 μm.
3. To add the scale bar to an image, select the image, and then
select “Analyze>Tools>Scale Bar.” The resulting dialogue
window (not shown) allows the user to position the scale bar,
set the scale bar length, and adjust the font size. The units are
the same as whatever units are entered in the Set Scale dialogue
window (Fig. 8f).
3.10 Generation of One useful operation that can be carried out with Fiji is the genera-
Heat Maps of Mineral tion of color-coded contour maps (heat maps) of mineral density
Density which are effective at illustrating quantified data. A typical example
is shown in Fig. 9 where the grayscale pixels comprising a longitu-
dinal section through a rodent incisor have been colored depending
on their grayscale value. This result was achieved using the “Inter-
active 3D Surface Plot” plugin written by Kai Uwe Barthel. This
plugin is included with ImageJ but needs to be installed when using
Fiji. This is simply a matter of going to https://imagej.nih.gov/ij/
plugins/surface-plot-3d.html and downloading the necessary file
Fig. 9 Using the “Interactive 3D Surface Plot” plugin to generate false color maps of mineral density. The
“Interactive 3D Surface Plot” plugin can assign different colors to different grayscale values (using a lookup
table LUT) which allow users to generate colored mineral maps which can visually emphasize changes in
mineral density. (a) A grayscale CT slice of a rodent mandible similar to the one shown in Fig. 2e was opened
and processed using “Plugins> Interactive 3D Surface Plot.” The plugin produces 3D models in which the
grayscale value of the pixels is represented in the Z-axis to give a height contour map of the mineral density.
To achieve the 2D result shown here, the Z-scale slider in the plugin window is set to zero and the image
rotated by right clicking with the mouse to remove any 3D effect. The first drop-down box is set to “Filled,” and
the LUT used here, “Fire,” is set using the second drop-down box. The calibration bar shows how the colors
correspond the original grayscale values of the pixels. (b) For presentation purposes, the scale bar can be
copied into PowerPoint, etc., and if the image has been calibrated against hydroxyapatite standards, the
grayscale values can be replaced with corresponding values for mineral density. Each grayscale value (0–255)
and its corresponding mineral density can be obtained by clicking the “List” button in the plot window shown
in Fig. 8c
from there. Once downloaded it can be installed by copying to the

plugins folder or by using “Plugins>Install plugin.” On restarting
Fiji, the plugin should appear in the list of available plugins
(“Plugins> Interactive 3D Surface Plot”). When the plugin is
applied to an image, a number of parameters can be adjusted,
such as the three-dimensional angle of view and how the grayscale
pixels are false colored (color is dictated by selecting a so-called
look-up table (LUT)). In general, the parameters shown in Fig. 9
give a satisfactory result. Be sure to select “Legend” under display
options as this shows the colored calibration bar and the grayscale
values corresponding to the colors (see Note 15). If mineral cali-
bration has been performed as described above, the mineral den-
sities associated with the grayscale values alongside the calibration
bar can be obtained by clicking the “List” button in the plot
window (Fig. 8c). The final image can then be calibrated in terms
of mineral density rather than just grayscale values as shown in
Fig. 9b.
Typically, the mineral density will be determined in multiple
slices to obtain mean data. It is possible to use the macro language
to automate this process and measure the mineral density of enamel
present on hundreds of slices in a whole image stack with a single
mouse click. Using macros to analyze whole image stacks compris-
ing hundreds of slices is discussed in the next section.
3.11 Using Macros to So far we have largely considered the analysis of single CT slices.
Automate the Analysis However, there are occasions when we may wish to analyze whole
of Whole Image Stacks image stacks comprised of hundreds of CT image stacks, e.g., to
measure enamel volume or obtain the mean density of the whole
enamel rather than the density of a few representative slices.
With an image stack loaded in Fiji, the user can easily advance
to the next slice to effectively section their way through the speci-
men. Each two-dimensional image comprising the stack is built up
of square pixels. Each pixel can be regarded as the front face of a
cube or voxel whose depth represents the apparent thickness of
each slice. The volume of each voxel is equal to the image pixel size
cubed. Measuring the area of enamel in each slice in terms of the
pixel gives the number of voxels in that slice. Summing the areas of
all slices gives the volume of the enamel in terms of the number of
voxels present. Multiplying this number by the volume of a single
voxel therefore gives an estimate of the enamel volume.
Once the enamel has been segmented by thresholding and a
ROI generated, we have already seen that it is a simple task to
determine the area of enamel bounded by the ROI. However, to
manually carry out this process on hundreds of consecutive slices in
an image stack representing the whole enamel is extremely labor
intensive. Fortunately, the process can be automated using the
ImageJ macro language.
In the example below, a macro has been written to measure the

area of mouse molar enamel on every slice comprising the image
stack.
1. The macro uses the “ROI manager” which simplifies handling
the hundreds of ROIs generated when analyzing whole image
stacks.
2. In essence, the user first thresholds the enamel, and the macro
despeckles each slice and generates an ROI around the enamel.
3. The ROI is added to the ROI Manager which allows for ROIs
to be overlaid on each slice so the quality of thresholding can be
assessed and in extremis allows each ROI to be manually cor-
rected by the user if required.
4. The macro generates a Results window giving the area and
density of enamel from each slice is the stack.
5. This data can be copied into Excel, etc. for further analysis.
6. Finally, the ROI Manager can save ROIs used to carry out the
analysis which simplifies revisiting the data at a later date.
As can be seen below, the macro is slightly more sophisticated
than the one described earlier in that it uses looping statements to
process multiple slices in image stacks, defines variables, uses con-
ditional statements (if/else) during its execution, and provides
prompts for the user.
//Macro 2
name=getTitle;
run("Threshold. . .");
msga = "Set threshold values then click OK ";
waitForUser(msga);
run("Analyze Particles. . .", "size=50-infinity show=Masks stack");
rename("Mask");
n = getSliceNumber();
while (getSliceNumber()<nSlices){
type = selectionType();
if (type==-1)
run("Next Slice [>]");
else
roiManager("Add");
run("Next Slice [>]");}
type = selectionType();
if (type==-1)
run("Next Slice [>]");
else
roiManager("Add");
selectWindow("Mask");
close();
selectWindow(name);
resetThreshold();
msga = "Scroll through ROI manager to check segmentation of enamel. Click
OK to see results.";
waitForUser(msga);
roiManager("Deselect");
run("Set Measurements. . .", "area mean standard redirect=None decimal=1");
roiManager("Measure");
1. To use the macro, open the stack to be analyzed, and then

select “Plugins>Macros>Run” to select the macro file (saved
somewhere on the system as a txt. file). Alternately, the macro
can be installed as described previously.
2. The macro duplicates the image stack and renames it Stack,
runs the Threshold applet, and generates a message and OK
button asking the user to set the threshold.
3. The user manually sets the threshold with the sliders or uses
one of the auto-thresholding algorithms in the drop-
down menu.
4. The user can scroll through the stack to check each slice is
thresholded satisfactorily (the macro will remove any speckling
later) and clicks OK.
5. The macro then runs “Analyze Particles” and removes any
thresholded object smaller than 50 pixels from each slice (this
value can be edited to optimize the macro performance
depending on the image stack in question).
6. Next the macro creates an ROI on slice one by creating a
selection around thresholded objects bigger than 50 pixels
and lists this ROI in the “ROI manager.”
7. If there are no objects to select, the macro advances to the next
slice until every slice has been processed.
8. The ROI Manager lists all slices where enamel has been seg-
mented, and clicking on a slice in the ROI Manager allows the
user to assess the quality of the thresholding achieved as the
ROI is superimposed on the relevant slice in the image stack
(the up and down keys on the keyboard can be used to scroll
through the whole list).
9. As with any selection generated in Fiji, the “Selection Brush”
tool (accessed by right clicking in the right-hand corner of the
“Oval Selection” tool button) can be used to either delete any
unwanted speckling or even adjust the shape of the ROI deli-
neating the enamel. Any changes made using the “Selection
Brush” tool can be saved by clicking “Update” in the ROI

Manager window.
10. Clicking OK generates the “Results” window showing the area
of enamel in each slice and the mean grayscale value with
associated standard deviations (see Note 16).
Greater sophistication could be achieved by using a more pow-
erful thresholding routine such as the “Trainable Weka Segmenta-
tion” plugin described earlier, but the simple example shown here
serves to show how macros can be used to greatly reduce the
workload involved in analyzing image stacks containing hundreds
of slices.
4 Notes
1. If the folder contains any files that are not part of the stack, use
the dialogue box to exclude these, or simply remove them from
the folder prior to importing as Fiji expects all images compris-
ing a stack to be of the same size.
2. The Reslice dialogue box allows the user to choose the number
of slices that will comprise the orthogonal stack (when the line
is drawn from the top of the window downward, the stack will
be resliced to the left of the line—when the line is drawn
upward from the bottom of the window, the stack is resliced
to the right of the line). If a rectangular box is drawn on the
image using the “Rectangular Selection Tool,” then “Ima-
ges>Stacks>Reslice” command can be used to generate an
orthogonal stack reslicing from the top, bottom, or either
side of the rectangle.
3. As described later, the grayscale values can be converted to
actual mineral density values if suitable hydroxyapatite calibra-
tion standards are available.
4. The problem is how does one select or segment the enamel
while leaving the other tissues unselected? The simplest way to
create an ROI is to draw it free hand using one of the drawing
tools.
5. It is obvious that defining ROIs by hand is time consuming
and, more importantly, increases the chances of operator bias
and inter- and intra-operator variability.
6. In attempting to threshold the enamel, some pixels in the
dentine and mandibular bone have also been selected.
7. This makes it easier to see exactly what has been selected.
8. Generating ROIs using thresholding can be more consistent
than drawing by hand as predefined upper and lower threshold
values can be applied to every slice, so the criteria used to define
the ROI are constant. However, the lower and upper threshold
values are still determined by the user and are therefore poten-
tially susceptible to operator bias. Ideally, it would be preferable
to use an auto-thresholding method that completely removes
the need for any input from the user and thus further reduces
operator bias and variability.
9. Exactly how these algorithms analyze pixel data is beyond the
scope of this chapter, but they exploit statistical methods and
fuzzy set theory to determine the similarity existing between
pixels in an image. More information on this subject can be
found at https://imagej.net/Auto_Threshold.
10. Documentation describing the ImageJ macro language is avail-
able on the NIH web site at https://imagej.nih.gov/ij/devel
oper/macro/macros.html, and numerous examples of macros
are available on the web, and users can quickly develop their
own macros by following the many examples already available.
11. Macros are discussed again later in the chapter when they are
used to automate the processing of whole image stacks rather
than single images as described above.
12. We do not need to accurately delineate these different areas; we
just need to provide examples of the pixels that comprise these
areas.
13. Out of necessity, this is only a brief introduction to using the
Trainable Weka Segmentation tool, and we have not men-
tioned the numerous settings that can be modified to optimize
the performance of the tool, but readers are referred to http://
imagej.net/Trainable_Weka_Segmentation for more details.
14. Note carrying out the measurement with all three enamel ROIs
selected as shown in the figure will return a single value which is
the mean of the three ROIs.
15. For presentation or publication purposes, this image can be
saved as an image file and edited in PowerPoint, etc.
16. The macro text is available online together with a video
showing the macro in use. Macros are easily adapted; for exam-
ple, adding the line run("Gaussian Blur. . ."); at the start of the
macro will run the Gaussian blur filter applet, so smoothing can
be applied to the image stack before the macro attempts to
carry out thresholding.
References
1. Brookes SJ, Barron MJ, Boot-Handford R, 2. Brookes SJ, Barron MJ, Smith CEL, Poulter
Kirkham J, Dixon MJ (2014) Endoplasmic JA, Mighell AJ, Inglehearn CF, Brown CJ,
reticulum stress in amelogenesis imperfecta Rodd H, Kirkham J, Dixon MJ (2017) Amelo-
and phenotypic rescue using genesis imperfecta caused by N-terminal enam-
4-phenylbutyrate. Hum Mol Genet elin point mutations in mice and men is driven
23:2468–2480
by endoplasmic reticulum stress. Hum Mol 8. Schindelin J, Arganda-Carreras I, Frise E,

Genet 26:1863–1876 Kaynig V, Longair M, Pietzsch T, Preibisch S,
3. Cao Z, Jiang B, Xie Y, Liu CJ, Feng JQ (2010) Rueden C, Saalfeld S, Schmid B, Tinevez JY,
GEP, a local growth factor, is critical for odon- White DJ, Hartenstein V, Eliceiri K,
togenesis and amelogenesis. Int J Biol Sci Tomancak P, Cardona A (2012) Fiji: an open-
6:719–729 source platform for biological-image analysis.
4. Feng J, Yang G, Yuan G, Gluhak-Heinrich J, Nat Methods 9:676–682
Yang W, Wang L, Chen Z, Schulze McDaniel J, 9. Arganda-Carreras I, Kaynig V, Rueden C, Eli-
Donly KJ, Harris SE, MacDougall M, Chen S ceiri KW, Schindelin J, Cardona A, Sebastian
(2011) Abnormalities in the enamel in bmp2- Seung H (2017) Trainable Weka Segmenta-
deficient mice. Cells Tissues Organs tion: a machine learning tool for microscopy
194:216–221 pixel classification. Bioinformatics
5. Pugach MK, Gibson CW (2014) Analysis of 33:2424–2426
enamel development using murine model sys- 10. Hall M, Frank E, Holmes G, Pfahringer B,
tems: approaches and limitations. Front Reutemann P, Witten IH (2009) The WEKA
Physiol 5:313 data mining software: an update. SIGKDD
6. Schmitz JE, Teepe JD, Hu Y, Smith CE, Explor Newsl 11:10–18
Fajardo RJ, Chun YH (2014) Estimating min- 11. Smith CE, Murillo G, Brookes SJ, Poulter JA,
eral changes in enamel formation by ashing/ Silva S, Kirkham J, Inglehearn CF, Mighell AJ
BSE and microCT. J Dent Res 93:256–262 (2016) Deletion of amelotin exons 3-6 is asso-
7. Schneider CA, Rasband WS, Eliceiri KW ciated with amelogenesis imperfecta. Hum Mol
(2012) NIH Image to ImageJ: 25 years of Genet 25:3578–3587
image analysis. Nat Methods 9:671–675
Chapter 27
Scanning Electron Microscopy (SEM) Methods

for Dental Enamel
Steinar Risnes, Muhammad Saeed, and Amer Sehic
Abstract
Scanning electron microscopy (SEM) is exceptionally well suited for the study of the structure of dental
enamel, due to its ability to create high-resolution images of hard surfaces. Continuous attention on how to
arrive at the observation stage with a clean and dry specimen is one main aspect of specimen preparation.
Other main aspects are choice of whether the specimen should be embedded or not, choice of plane of
section, and choice of acid-etching regime. Special attention is given to the preparation of small specimens
and how to prepare and observe more than one plane or aspect in the same specimen.
Key words Scanning electron microscopy (SEM), Dental enamel, Teeth
1 Introduction
Dental enamel is the hardest tissue in the body. It consists mainly of

the mineral hydroxyapatite, making up about 95% per weight and
87% per volume of the enamel. The mineral is present in the form of
small crystals, tightly packed. The crystals are not randomly
distributed. They form a repetitive pattern which serves two main
purposes: hardness and strength. Hardness since the pattern allows
tight packing of crystals and strength because the crystals are dis-
posed in two main and different directions, in prisms and inter-
prism, thereby reducing the risk of crack formation. Also, in this
pattern, when properly interpreted, lies information about enamel
formation: ameloblast size, packing, movement, rate of secretion,
rate of extension/differentiation, and life span.
SEM is mainly a method for the study of morphology and
structure, at a wide range of magnifications. Images formed by
detecting secondary electrons primarily provide information
about the topography of the surface that is scanned by the electron
293
294 Steinar Risnes et al.
Fig. 1 Schematical outline of main steps in preparation of enamel specimens for SEM (see text)
beam. Applied to dental enamel, two main aspects emerge: the

study of enamel surface and the study of internal enamel structure.
The latter can be achieved by studying etched sectioned surfaces
and by studying unetched fractured surfaces. Backscattered elec-
trons can give some information about the immediate subsurface
structure, but this aspect will not be dealt with here.
The enamel surface proper is the part of the enamel that inter-
acts with the oral environment. It is subjected to wear, both
mechanical by abrasion and attrition and chemical by erosion and
caries. In addition, organic and/or inorganic material may be
deposited, e.g., plaque, pigments, and calculus. The present proto-
col deals with the enamel tissue itself, not with material deposited
on its surface.
Various aspects of materials and methods involved in the pres-
ent methodological approach to SEM studies of dental enamel are
presented schematically in Figs. 1, 2, and 3.
SEM Methods for Dental Enamel 295
Fig. 2 Schematic representation of method and equipment for preparation of small specimens. In (A) is shown
a small cylindrical brass specimen stub on which a small specimen is glued and which fits in a specially
designed holder. In (B) is shown the assembly shown at the bottom of panel a placed in the adjustable cylinder
of the sectioning machine’s specimen holder (Fig. 2). Precision grinding of side aspects can now be performed
under a dissecting microscope using a simple grinding paper assembly. In (C) the specimen stub with
mounted specimen is positioned horizontally in the stub holder and put in the SEM, allowing observation of
several aspects by rotating the cylindrical specimen stub in the holder
2 Materials
2.1 Teeth (See

Note 1)
2.2 Resin (“Epon” 1. Preparation of “Epon” (Epoxy Embedding Medium, Sigma-
and Spurr) Aldrich/Fluka): Mix 49 g DDSA, 49 g MNA, and 103 g
“Epon” and stir for half an hour. Add 2 g DMP-30 and stir
for 15 min. The mixture can be stored in plastic syringes in a
freezer ( 18 C) for about 6 months. Curing regime: room
temperature for 1 day, 45 C for 1 day, 60 C for 2 days, and
room temperature for 1 day (see Note 2).
Fig. 3 Schematic representation of sectioning machine. The machine’s speci-

men holder is adjustable (rotation), allowing specimens to be sectioned along
predetermined planes. The specimen holder’s surface is partly grooved to
facilitate wax retention. An adjustable cylinder is used for holding small speci-
mens when the holder is used for precision grinding (see Fig. 2)
2. Preparation of Spurr (Agar Scientific): Mix 10 g ERL 4206, 8 g

DER 736, 26 g NSA, and 0.4 g S 1, and stir carefully. The
mixture has a pot life for 3–4 days. Cure at 70 C for 9 h (see
Note 3).
3. Resin embedding molds of various sizes.
2.3 Wax 1. Sheets of Tenax wax (SS White).

2. Spatula/knife for warming/melting wax.
2.4 Diamond Blade 1. Diamond wafering blade (10.2 cm 0.3 mm) for sectioning
machine (Buehler) (Fig. 3).
2.5 Grinding Paper 1. Sheets of waterproof grinding paper, grits 1200 and 600 (3 M).
and Powder 2. Standard histological glass slides (used with grinding paper for
grinding thin sections).
3. Micropolish 0.3 μm alpha alumina powder (Buehler).
2.6 Specimen Stubs 1. Standard aluminum SEM stubs (Fig. 1).

2. Brass cylinders, 7 4 mm, obtained by cutting 7 mm-long
pieces off a 4 mm-diameter brass rod (Fig. 2).
2.7 Glue 1. Cyanoacrylate glue.

2. Adhesive carbon tape for standard specimen stubs.
3. Silver paste.
2.8 Brushes 1. Artists’ brushes and toothbrushes of various degrees of stiffness

and density.
2. Soap water (dish washing concentration) for brushing/
cleaning.
3. Absorbent paper for drying.
2.9 Acids 1. Nitric acid of various concentrations, e.g., 0.1%, 0.5%, 1%, and
2.5%. Prepare with distilled water.
3 Methods
Careful planning of how to bring a specimen of dental enamel

through a preparational process to a point where it can be subjected
to a meaningful observation in the SEM includes initial scrutiny of
specimen and awareness of subsequent procedures. These may
include embedding in resin, sectioning, gluing to specimen stubs,
grinding/polishing, cleaning, etching, drying, and sputter coating.
It is often advantageous and/or necessary to reduce the size of a
tooth by sectioning, retaining the part that is of interest for study in
the SEM.
3.1 Planning 1. Observe specimen in dissecting microscope (see Note 4).

2. Small specimens may be considered for embedding in a resin
(see Note 5).
3. If the whole tooth is to be observed in the SEM, go to Sub-
heading 3.4.
4. Determine and mark plane(s) of section (see Note 6).
3.2 Sectioning 1. Fix specimen with Tenax wax to specimen holder in sectioning
machine (Fig. 3) (see Note 7).
2. Section specimen in sectioning machine using ample water for
cooling (see Note 8).
3.3 Grinding and 1. If sectioned surface is not to be observed in SEM, go to

Polishing Subheading 3.4.
2. Grind and polish the sectioned surface that is to be observed in
SEM (see Note 9).
3.4 Gluing Specimen 1. Clean the surface that is to be glued to specimen stub (see Note
to Specimen Stub 10).
2. Dry the specimen (see Note 11).
3. Glue specimen to stub (see Note 12).
3.5 Cleaning 1. Brush with soap water the surface that is to be observed in SEM
Specimen (see Note 13).
2. Ultrasonicate in distilled water (see Note 14).
3.6 Acid Etching of 1. Specimens are etched by dipping them in an acid for an ade-
Specimens quate period of time (see Note 15).
3.7 Fracturing 1. Specimens are fractured with chisel and hammer (see Note 16).
Specimens
3.8 Drying Specimen 1. Specimens are dried before sputter coating and observation in
the SEM (see Note 17).
3.9 Securing 1. Silver paste may be added to secure electrical conductivity of

Electrical Conductivity the specimen (see Note 18).
of Specimen 2. Sputter coating (see Note 19).
3.10 Observation 1. (See Note 20).

in SEM
4 Notes
1. Teeth may be fresh (newly extracted), fixed, or dried. Newly

extracted teeth should be cleaned/brushed thoroughly and
stored in 70% ethanol. Fresh and fixed teeth should be sub-
jected to gentle dehydration in a graded series of ethanol
(30 min in each of 50, 60, 70, 80, 90, and 100%). The surface
of teeth that has been fixed in ethanol, formalin, or glutaralde-
hyde may be more difficult to clean due to hardening/conden-
sation of organic material on the surface.
2. The ratio of DDSA and MNA determines block hardness,
increasing MNA gives harder blocks.
3. A rapid cure may be effected in 3 h by increasing the quantity of
S 1 to 1.0 g, but this reduces pot life to 2 days. Harder blocks
are obtained by reducing the quantity of DER 736.
4. When observing the specimen in a dissecting microscope [1],
play with the direction of illumination. In already sectioned
specimens, especially longitudinally sectioned teeth, prism ori-
entation is usually evident. Sectioned specimens may be put in a
Petri dish with water and illuminated directly from above or

indirectly from below, enhancing transparency and visibility of
three-dimensional features.
5. Some specimens may be judged so small that it is best to embed
them in a resin, preferably Epon or Spurr [2], in order to
facilitate subsequent handling [3, 4].
6. Marking position and orientation of planes for subsequent
sectioning can be made at this point, using a fine pencil or a
sharp instrument. A small specimen embedded in a resin is
usually readily visible through the transparent resin. To increase
the visibility of an embedded specimen, it can be helpful to
reduce the size of the resin block and to polish its various
aspects. Sometimes it is advantageous to observe the enamel
structure in different planes cut through the same specimen
[5–7]. For human teeth this is most easily done by first obtain-
ing a relatively thick facio-lingually cut longitudinal section,
and on this mark a second plane, e.g., transverse to the prisms,
and even a third plane parallel with the prisms. Multiplane
sectioning, grinding/polishing, and observation in SEM, espe-
cially of small specimens, are facilitated by the use of a simple,
homemade (university’s workshop) apparatus [8] (Fig. 2).
7. For sectioning, we use a homemade (university’s workshop)
version of Gillings–Hamco thin-sectioning machine [9]
equipped with an adjustable specimen holder [10] (Fig. 3).
The specimen holder sits on a table which can be moved
manually forward, backward, and sideways, as well as with a
slow, adjustable, motor-driven movement forward toward the
machine’s rotating diamond blade. The diamond blade will just
go clear of the holder’s top surface during sectioning. The
holder’s grooved top surface should have a thin layer of
Tenax wax whereupon teeth/specimens may be fixed by soft-
ening the wax with a hot instrument and melting and applying
additional wax for additional support. Take care not to con-
taminate with wax the surface that is intended for observation
in the SEM. This is especially important when the enamel
surface is to be observed, but less critical for sectioned planes
since these are going to be ground and polished subsequently.
The specimen holder can be rotated, thereby allowing adjust-
ment of the plane of section relative to the plane of the dia-
mond blade. Also, the whole table of the machine can be
moved sideways, facilitating correct positioning of the speci-
men relative to the diamond blade.
8. Use very slow specimen feed toward the rotating diamond
blade and ample water for cooling. Since enamel is the tissue
of interest, the root may be cut off (Fig. 1A). This can be
achieved by rotating the specimen holder so that the longitu-

dinal axis of the tooth is oriented perpendicular to the diamond
wheel (Fig. 3). If the enamel surface is the area of interest, the
whole crown can be used after cutting off the root (Fig. 1B), or
the crown may be sectioned longitudinally (Fig. 1), either
mesio-distally or facio-lingually, allowing observation of facial
and lingual or mesial and distal aspects, respectively (Fig. 1E, F).
When enamel is sectioned for study of enamel structure, two
parallel sections should be made (Fig. 1C), one for the surface
intended for study and the other for gluing to the stub, yielding
a stable and adequate position of the specimen with an observ-
able surface perpendicular to the electron beam (Fig. 1F). Dur-
ing sectioning the moving table is covered with a transparent
plastic hood in order to minimize water spill.
9. A sectioned surface intended for the study of enamel structure
should be ground and polished in order to get rid of the
grooves and scratches caused by the diamond blade during
sectioning (Fig. 1D). We use grit 1200 3M® waterproof grind-
ing paper for grinding, and for polishing we use its back side
with water and 0.3 μm alumina powder. A relatively thick
section can easily be handled using finger pressure (Fig. 1D).
Thinner sections may be ground/polished using the method of
Frost [11] with the specimen supported by a coarse grinding
paper wrapped around a glass slide and ground against a finer
grit grinding paper or polishing surface. Grinding/polishing of
small specimens using equipment dedicated for that purpose
(Fig. 2) [8] is best performed under a dissecting microscope.
10. The specimen surface that is to be glued to the specimen stub
should be cleaned shortly by brushing with soap water and
rinsed under running tap water.
11. Specimens need to be dry before observation in the SEM. They
also need to be dry before the final step of sputter coating.
Indeed, residual moisture is removed during evacuation of the
sputter chamber, but most of it should be removed before that
step. Dentine holds more water than enamel and some dentine
will be present in most enamel specimens. It takes time for
moisture to escape from both these mineralized tissues. In the
next step of specimen preparation (see Note 12), the dentine
aspect of the specimen will be glued to the specimen stub using
a cyanoacrylate glue. This seals off an important escape route
for moisture. Therefore, it is suggested that specimens are
dried before gluing (Fig. 1B, D). If the tooth initially was dry
or had been ethanol-stored or ethanol-dehydrated, drying at
room temperature (25 C) overnight may suffice. If the tooth
was fresh/wet/fixed and had not been stored in ethanol or
subjected to ethanol dehydration, a longer drying period may
be needed, for instance 2–3 days. After gluing (next step), an
important moisture entrance route will have been blocked, and

this will probably retard remoistening on subsequent contact
with water.
12. If the occlusal surface of a molar or a premolar is the aspect of
interest, the cutoff crown, after cleaning and air-drying, is
glued to the stub, using a cyanoacrylate glue and a standard
aluminum stub (Fig. 1B). The same applies to specimens sec-
tioned for observation of crown side aspects (Fig. 1E). Gluing
the specimen to the stub with a cyanoacrylate glue yields a
strong bond that withstands and facilitates subsequent
handling. The stub pin can subsequently be used as a handle.
Do not use an excess of glue; it should preferably be restricted
to the area of contact between specimen and stub and just a
little beyond. Some extra glue may be used to carefully seal off
additionally exposed dentinal surfaces. If the specimen will not
be subjected to further mechanical or chemical treatment, it
may be fixed to a standard SEM stub using adhesive
carbon tape.
13. Since specimens probably have been touched by fingers during
prior handling, and may have been soiled by other sources as
well, it should be cleaned by brushing in soap water for at least
1 minute, using medium-to-stiff and dense artist’s brushes
(Fig. 1b, e); neither soap nor brush affects the hard, inorganic
enamel surface, with one exception: in areas of demineraliza-
tion, some material may be removed by the brushing. After
cleaning, rinse under running tap water, with hard water rush
and long working distance.
14. Put specimens in a jar with distilled water in an ultrasonication
basin for about 5 min. Repeat with refreshed distilled water.
Remove excess water with absorbent paper, avoiding direct
contact with the surface that is to be observed in the SEM.
15. In order to reveal the structure of enamel in sectioned planes,
the specimen can be etched by acid. Since prism and interprism
crystals are differently oriented and since enamel crystals are
attacked differently depending on their orientation relative to
the surface where the acid is applied [12], acid etching will
always yield a structure-dependent surface topography inde-
pendent of orientation of the sectioned surface. The thickness
of the enamel removed by etching increases with acid concen-
tration and etching time [13], and, up to a point, so does the
degree of topography/roughness [14]. Various acids and
EDTA have been used [15, 16]. We prefer etching with nitric
acid of different concentrations, from 0.1% to 5%, mostly in the
lower range. Adequate etching regimes could be 0.1% nitric
acid for 30–60 s, 0.5% for 25–45 s, 1% for 15–25 s, and 2.5 and
5% for 10–15 s. Holding the specimen or the specimen stub
Fig. 4 Change in pH with time when storing 1, 2, and 3 human third molars in
90 mL of 1% nitric acid
with a pair of tweezers, dip it into a jar/container with acid,

and move the specimen quite vigorously during etching in
order to prevent reprecipitation of dissolved material. One
should only etch so much as to disclose the enamel structure,
main features being prisms, interprism, prism sheaths, Retzius
lines, prism cross striations, and Hunter–Schreger bands.
Fewer Retzius lines are observable in the SEM compared to
light microscopy of ground sections, probably since the latter
gives a summation image from the whole thickness of the
section, while the former only represents a single plane. The
more brutal the etch, the more difficult it is to interpret the
structure. This has to do with unduly accentuated topography
and widened prism sheaths. Acids should be changed often in
order to retain an adequate ionic strength and pH. With time
etching of teeth increases the pH in a limited volume of acid
(Fig. 4). Specimens should be rinsed in running distilled water
immediately after etching in order to stop the etching process
and remove residual acid from the specimen.
16. Fracturing is an alternative method for studying internal
enamel structure. It is easily performed using a chisel and a
hammer. The disadvantage of the method is that the position
and orientation of the fracture plane is unpredictable and the
disclosed structure may be difficult to interpret. However, it
has the advantage over acid etching that no artifacts are
introduced, other than the fracture itself. The method can be

beneficial for studying fine details of enamel crystals and their
organization. It may also demonstrate prism paths in three
dimensions since fractures occasionally occur along the prisms.
The fractured surface is pristine and needs only to be subjected
to light brushing, rinsing, and ultrasonication in order to get
rid of loose fragments.
17. A final drying before sputter coating could be at room temper-
ature for 2–3 days (Fig. 1f). A graded series of alcohol should
not be used at this point since this may weaken the cyanoacry-
late glue and also contaminate the specimen. As a final and
extra drying measure, the specimen may be placed in a desicca-
tor for about 2–3 h. Drying at higher temperature in a drying/
warming cabinet reduces the drying time, but increases the risk
of crack formation. In general, the larger the specimen, the
longer the drying time.
18. Silver paint may be applied at the specimen stub junction,
especially for bulky specimens like human teeth. It may be
extended to those parts of the specimen that are not of interest
for observation. The silver paint will, together with the subse-
quently sputtered metal layer, provide electrical conductivity
preventing charging of the specimen. Do not use excess silver
paint; it is more its distribution than its mass that is important.
19. Sputter coating with a metal layer. This is in order to provide a
layer through which static electricity may slip away during
observation when the specimen is scanned with a focused
electron beam. We routinely apply a 30-nm-thick layer of
gold palladium, a compromise giving adequate conductivity
and minimal/acceptable veiling of fine details. It is possible
to observe uncoated enamel specimens in the SEM [17], but
the accelerating voltage has to be reduced (3–5 kV), which will
also reduce resolution. Using uncoated specimens opens the
possibility of performing serial etchings.
20. Observation in SEM. We use acceleration voltages between
12 and 25 kV, more often in the lower range. When increasing
the working distance, the depth of focus will increase, which is
advantageous when observing an uneven or oblique surface.
When observing a flat surface perpendicular to the beam, a
shorter working distance can be used, giving an improved
resolution. Observations of the enamel surface, intact and
etched, are shown in Fig. 5. Observations of incremental
lines, some in several planes, are shown in Fig. 6. Observations
of the effect of various acid-etching regimes are shown in
Fig. 7. Observations of fractured enamel are shown in Fig. 8.
Fig. 5 SEM micrographs of intact (a–c) and acid-etched (d–f) enamel surface. Perikymata are evident in (a),
prism profiles (P) of variable distinctness are discernible in (b and c). Prisms are much more obvious in enamel
etched for 1 min in 0.5% nitric acid (d–f), except in areas mixed with prism-free enamel (PFE). The etching has
not completely removed scratches on the surface (d). Bars are 200 μm for (a and d); 20 μm for (b, c, e, and f)
Fig. 6 SEM micrographs of human enamel etched with various concentrations of nitric acid for various periods
of time. (a–c) Shows increasing distinctness of prisms and interprism but also increasing distortion of the
enamel structure. (d–f) Shows decreasing distinctness of prisms and interprism and some additional distortion
of the enamel structure. P prism, IP interprism. Bar is 5 μm
Fig. 7 SEM micrographs of acid-etched human (a–d) and mouse (e) enamel. (a, b) Observation of enamel in
different sectioned/ground planes in the same specimen, demonstrating continuity of Retzius lines (arrows)
across planes which are at approximately right angles to each other: longitudinal planes (L), transverse planes
(T, fractured in (a)), and tangential plane (Ta). (c) Outer enamel with distinct Retzius line (arrow) showing
altered distribution of prism and interprism domains and prism with distinct cross striations (arrowheads). (d)
Hypomineralized enamel. Arrow indicates prism direction. Prisms exhibit distinct cross striations. (e) Cervical
enamel of mouse mandibular first molar showing two to three incremental lines in the outer enamel (arrows).
Fig. 8 Fractured human enamel. (a) Typical longitudinal fracture surface with a distinct zone of Hunter–S-
chreger bands (HS). D dentin. Bar is 500 μm. (b–d) Fracture surfaces after different procedures; (b) specimen
was not brushed or rinsed; numerous enamel fragments are present; (c) specimens were brushed under
running tap water; fewer fragments remain; (d) specimen was rinsed under running tap water and ultra-
sonicated for 5 min; some rounded grains remain. P prism, bars are 20 μm in (b and c), 10 μm in (d)
References
1. Risnes S (1973) Three-dimensional features of 3. Cuny G, Risnes S (2005) The enameloid
human enamel as seen with the dissecting microstructure of the teeth of synechodonti-
microscope. Arch Oral Biol 18:647–650 form sharks (Chondrichthyes, Neoselachii).
2. Li C, Risnses S (2004) A comparison of resins Palarch’s J Vertebr Palaeontol 3:8–19
for embedding teeth, with special emphasis on 4. Sehic A, Peterkova R, Lesot H, Risnes S (2009)
adaptation to enamel surface as evaluated by Distribution and structure of the initial dental
scanning electron microscopy. Arch Oral Biol enamel formed in incisors of young wild-type
49:77–83 and Tabby mice. Eur J Oral Sci 117:644–654
Fig. 7 (continued) (a, b, and d) have been etched with 1% nitric acid for 15 s, (c) in 0.1% nitric acid for 30 s,
and (e) in 0.1% nitric acid for 45 s. P prism, D dentin, R embedding resin. Bars are 100 μm in (a), 50 μm in (b),
20 μm in (c), 10 μm in (d), and 20 μm in (e)
5. Risnes S (1985) A scanning electron micro- 12. Simmelink JW, Nygaard VK, Scott DB (1974)
scope study of the three-dimensional extent of Theory for the sequence of human and rat
Retzius lines in human dental enamel. Scand J enamel dissolution by acid and by EDTA. A
Dent Res 93:145–152 correlated scanning and transmission electron
6. Risnes S (1987) Multiplane sectioning and microscope study. Arch Oral Biol 19:183–197
scanning electron microscopy as a method for 13. Fosse G (1968) A quantitative analysis of the
studying the three-dimensional structure of numerical density and the distributional pat-
mature dental enamel. Scanning Microsc tern of prisms and ameloblasts in dental enamel
1:1893–1902 and tooth germs. II. Serial etching of dental
7. Risnes S (1990) Structural characteristics of enamel. Acta Odontol Scand 26:285–314
staircase-type Retzius lines in human dental 14. Sehic A, Risnes S, Khuu C, Khan Q-E-S,
enamel analyzed by scanning electron micros- Osmundsen H (2011) Effects of in vivo trans-
copy. Anat Rec 226:135–146 fection with anti-miR-214 on gene expression
8. Risnes S (1985) Multiangular viewing of dental in murine molar tooth germ. Physiol Genomics
enamel in the SEM. An apparatus for con- 43:488–498
trolled mechanical specimen preparation. 15. Poole DFG, Johnson NW (1967) The effects
Scand J Dent Res 93:135–138 of different demineralizing agents on human
9. Gillings B, Buonocore M (1959) An apparatus enamel surfaces studied by scanning electron
for the preparation of thin serial sections of microscopy. Arch Oral Biol 12:1621–1634
undecalcified tissues. J Dent Res 16. Boyde A, Jones SJ, Reynolds PS (1978) Quan-
38:1156–1165 titative and qualitative studies of enamel etch-
10. Risnes S (1981) A rotating specimen holder for ing with acid and EDTA. Scanning Electron
hard tissue sectioning. Stain Technol Microsc 2:991–1002
56:265–266 17. Risnes S, Stølen SO (1981) Uncoated speci-
11. Frost HM (1958) Preparation of thin undecal- mens of human enamel observed in the scan-
cified bone sections by rapid manual method. ning electron microscope. Scand J Dent Res
Stain Technol 33:273–277 89:205–212
Chapter 28
Microcomputed Tomography Imaging in Odontogenesis

Studies
Kostas Verdelis and Phil Salmon
Abstract
3D analysis of animal or human whole teeth and alveolar bone can be performed with high sensitivity in a
nondestructive manner by microcomputed tomography. Here we describe the protocols to be followed for
the most common applications in the developmental studies of dental and craniofacial tissues. Emphasis is
placed on the basis of choosing settings for image acquisition, such as voxel resolution (Fig. 1), or beam
energy (Fig. 2) and for processing, such as segmentation method (Fig. 3), parameters. The limitations to
take into account for optimal efficiency and image quality are also explained.
Key words Microcomputed tomography, Dentin, Enamel, Dental tissues, 3D analysis, Beam harden-
ing, Beam energy, Resolution
1 Introduction
Microcomputed tomography (microCT) was developed in the late

1980s with in-house-built units initially, was driven by bone char-
acterization and 3D morphometry analysis in the mid-1990s, and
has finally evolved into a ubiquitous analytical imaging method in a
wide range of fields, from biological sciences to bioengineering and
materials science [1]. It is based on the principle of acquisition of
continuous X-ray lateral projections (by an X-ray detector coupled
with a digital readout system) of a specimen rotating on a stage
while it is being irradiated by a conical beam from an X-ray source
with a micro-focal (to limit penumbral artifacts) spot as an output
[2]. The lateral projections are reconstructed into a 3D volume
(a file that contains the coordinates for all volume elements—vox-
els—inside the specimens, also containing an X-ray attenuation
coefficient value for each element) using back-projection algo-
rithms (Feldkamp Davis et al. 1984). Modern tabletop systems
can record architecture and density information of specimens
from a wide density range with a submicron theoretical (voxel)
309
310 Kostas Verdelis and Phil Salmon
resolution and a high sensitivity (described for dental hard tissues in

(Dong Dong et al. 2014)).
MicroCT started being extensively used in dental research
when high-resolution systems became commercially available in
the early 2000s, while currently numerous relevant reports exist
in almost every clinical and basic science area of the former
[3, 4]. The main advantages of the method that relate to dental
research are (1) 3D analysis (measurements from the whole volume
of interest, as well as directly derived measurements in space), which
provides statistically more robust data than the traditional 2D
analysis (on sections from or the surface of a specimen) modes,
(2) the possibility of digital reorientation of the imaged volumes for
an optimal viewing and processing (this reorientation is most often
based on internal landmarks, which is not possible in methods that
need beforehand preparation of the samples—see also Fig. 2), and
(3) the ability to use the same specimens for additional analyses
(histology, mechanical tests) due to the nondestructive nature of
the method. (1) and (2) are further explained in this chapter.
Although important data have been reported using newer
modes, such as nanoCT ([5, 6] and enhanced-contrast—based on
phase contrast optical principle—[7] microCT, this chapter will
refer to the conventional optic high-resolution systems as these
are used for the vast majority of applications. It finally has to be
noted that terminology for some microCT image acquisition and
3D reconstruction settings or analysis modes and parameters often
differs between manufacturers of systems and software. This is also
reflected in the literature, as different terms for the same parameter
tend to be used in reports where different systems/software have
been used. So, some of the terms used here may differ from the
respective ones the user is presented with when using a particular
system.
2 Materials
2.1 Preparation of 1. Specimens should generally be scanned inside a holder,

the Specimens although some manufacturers provide the option of scanning
a freestanding specimen. Every system provides a range of
holders usually made of a relatively radiolucent and chemically
resistant (e.g., polycarbonate) plastic to accommodate samples
of different size (or a single vs. multiple samples imaged in a
parallel batch mode).
2. The rule of thumb is that the narrowest in diameter holder that
can accommodate the specimen should always be used. This is
because a larger than needed holder increases the field of view
(FOV) required for a particular magnification unnecessarily.
Microcomputed Tomography Imaging in Odontogenesis Studies 311
3. The specimens should be stabilized with their long axis aligned

with this of the holder (scan’s z axis) to reduce beam hardening
(see Note 1). “Makeshift” holders inside the holder provided
by the manufacturer (e.g., a 15 mL Falcon tube containing the
specimens placed inside the standard holder) should be
avoided, as they affect the mineral density calibration.
4. The holder should be filled either with a liquid medium (saline,
deionized water, or 70% ethanol), or wrapped in moistened
paper tissue, depending on the nature of the samples and the
kind of further analysis planned, if any. (Moistened paper also
maintains hydration while adding less additional peripheral
xray absorption, thus improving image quality relative to scan-
ning in liquid). For instance, bones planned to undergo
mechanical testing should be kept in saline, or wrapped in
saline-wetted tissue or gauze, during imaging, whereas 70%
ethanol is acceptable if histological analysis is to follow.
5. Although scanning in air actually provides a sharper contrast
between the imaged object and the background, mineral den-
sity calibration is generally not acceptable without the sample
being fully hydrated, either in a liquid medium or wrapped in
moistened tissue. The option of “dry” scanning only exists
when the data expected are strictly visual and not numerical
(see Note 2).
6. It is best to physically dissect—using a scalpel blade or a drill
under the dissecting microscope—any tissue around the speci-
men that won’t be part of the analysis (e.g., the rest of a mouse
hemi mandible if only the molars are analyzed with a very high
resolution) if that does not compromise the specimen’s integ-
rity. Although it is not required, it can help in keeping the FOV
needed as small as possible for higher magnification without
increasing the scanning time, as well as reducing undue beam
hardening from hard tissue that is not part of the analysis.
7. A popular practice for imaging large numbers of specimens is to
pack them next to each other in a customised holder or “car-
ousel”, and scan all the samples mounted in this way at the
same time in a single scan. This significantly reduces the total
scan time needed for all samples. However there is an image
quality penalty for doing this: it results in a wider FOV neces-
sary to accommodate the packed specimens and thus resolution
is lower (pixel size is larger). Also, beam hardening artifacts are
increased. So, with this the parallel batch type of packing
samples for scanning, one should balance the advantages of
saved effort, time and/or scan cost, against the consequent
reduction in image quality.
8. Most microCT systems can accommodate a number of speci-
mens for sequential imaging in a vertical fashion (vertical batch
mode). The number of specimens stacked vertically does not

affect the quality of the scan. Care should be exercised to make
sure that consecutive specimens are separated from each other
by packing material or medium.
9. The specimens should be stabilized and separated from each
other to avoid a motion artifact or need for a tedious digital
separation of specimens in contact after image acquisition. The
most popular materials for the purpose have been Parafilm
sheets and gauzes, but any other relatively radiolucent and
not soluble in the medium used material may do.
10. Specimens scanned together in a parallel batch mode should be
packed in a way that enables identification in the reconstructed
volume. Toward that one can, e.g., record the order in which
specific specimens are placed clockwise, starting with one
packed with a small wax ball adjacent to it. The wax ball will
be visible in the reconstructed volume without interfering with
the analysis (wax is relatively radiolucent). The vertical order of
the specimens in a vertical batch should also be recorded.
11. Use of a contrast agent: Perfusion of soft or very young miner-
alizing tissues with radiopaque solutions, such as phospho-
tungstic acid (PTA), Lugol’s iodine (Iodine + KI), or HgCl2,
has been used with excellent results in very diverse applications
(from imaging of mouse arteries to biomaterials) to enhance
contrast [8]. Mineral density measurements cannot be per-
formed with this kind of imaging. Note that acid contrast
agents such as PTA will partly demineralise bone. Alternatively,
for imaging of young mineralized tissue, one can physically
dissect the object to be imaged from the surrounding hard or
soft tissue (e.g., the cervical part of a mouse incisor or uner-
upted mouse molars from the rest of the jaw) and use high-
contrast conditions (see below).
2.2 Setting Up 1. Selection of voxel size: As a general rule, the voxel size for the
the Scan imaging of an object should ideally be three or four times lower
than the thickness of the thinnest structures within the object
that are part of the measurements (Fig. 1). This is because the
voxel resolution (the voxel size that results from the particular
geometric—defined by the relative source-object/object-
detector distances—magnification and the detector binning
mode used) is always lower than the real resolution of the
resulting images (see Note 3). It is best to first do a pilot scan,
identify the thinnest structures of interest, and measure them
across using the standard measuring tool provided. However,
choosing the voxel size also has to take into account (1) the
available resulting FOV (the higher the voxel resolution, the
smaller the available FOV is) and (2) the associated scanning
time (and hence, cost). Most modern systems have two or
Fig. 1 Coronal views of an embryonic (day 16) mouse head volume. The location of the view plane is indicated
by the arrows in (a), (b) volume from a 2 μm voxel size scan, (c) volume from same scan as in B digitally
downgraded (binned) by a factor of 8 to a final voxel size of 16 μm, and (d) volume from a 10 μm voxel size
scan. Grayscale and binarized (segmented with the same threshold) views are shown in the respective images
of the upper and lower rows. The location of the incisor is shown by the dotted line in (b). Note that little detail
is lost from the original scan in the digitally resampled 16 μm volume, whereas the 10 μm volume suffers a
significant partial volume effect. This is obvious in the binarized views where small segments of the incisor
(such as the approximately 15 μm thick one pointed at by the arrow in b) are represented in the binarized view
in c but lost in the respective one in d. It is also evident that the image of the natural porosity of the young
incisor is retained in c but lost in d, as well as the rest of the incisor segments in d appear thicker in the
binarized views than they really are
more settings of the detector, corresponding to a factor of

binning (from 1 up to 4, see Note 4) of the detector elements.
The higher the binning factor for the same voxel resolution,
the more efficient the photon collection by the detector and the
shorter the scan is, albeit at the expense of a smaller available
FOV (it becomes half or one fourth the unbinned FOV size for
2 2 or 4 4 binning, respectively).
2. Some systems only provide an option for selection of specific
voxel sizes (e.g., 3, 6, 10, 15, 20, and 30 μm), as well as holders
of different diameters that correspond to each of the latter
(e.g., for 3 and 6 μm, for 10 and 20 μm, etc.). Other systems
provide the option to select any voxel size within an available
range (e.g., in the 6–14 μm or 3–7 μm range), while the user
has to adjust the holder size to the available FOV. In the second
case, it is very important to make sure that the whole holder
(and not only the specimen inside it) stays within the result-
ing—from the voxel size and detector binning mode selected—
FOV at all times during the rotation; otherwise partial volume

scan artifacts will be generated.
3. Selection of xray beam energy: Xray beam energy is set by the
combination of the xray filter and the applied voltage (Kvp).
Beam energy dictates the contrast within an imaged volume;
the lower the beam energy, the higher the contrast between
different tissues or a tissue and the background is. As a general
rule, the beam energy (filter and voltage) selected should be
the lowest for which there is at least 15% xray transmission at
the darkest part of the projection image, and also for which
beam hardening is not excessive (see Note 5) within the region
for analysis. For instance, if mature enamel is part of the region
of interest (ROI—the area in which measurements will be
performed) or right next to it, then a high beam energy setting
(at least 1mm Al filter and 70 Kvp, depending on the system
and source type) should be selected in most of the systems. If
the analysis is limited to only root dentin (specimens positioned
with long axis aligned to scan z axis so that the crown and root
are not in the same beam path), the range selected should be a
medium one (0.5–1 mm Al filter, 50–60 Kvp).
4. If mineral densities within the imaged object fall into a wide
range, an example is the continuously growing mouse incisor
with enamel areas of low and very high density [9] (Fig. 2); the
beam energy should be selected to match the range of mineral
densities within the ROI (see Note 6). For example, in the
continuously growing mouse incisor, such a high setting
(70–90 Kvp) should be selected if the incisal half—in which
mature enamel parts are included—is to be analyzed, but a
lower one (40–45 Kvp) is more suitable if areas close to the
cervical loop, young part of the incisor, are analyzed instead.
5. Some systems allow the user to select a specific setting within a
range of beam energies—for instance, any setting between
40 and 60 Kvp, as opposed to an exact 45 Kvp for high-
contrast conditions—while other systems only allow the selec-
tion of three or four Kvp values. The beam energy to be used is
a function of the factors discussed above, as well as whether the
scan is performed in a parallel batch mode.
6. Selection of other conditions: Most systems allow the user to
customize a scanning protocol by setting the exposure time
(time for which the detector is continuously irradiated at every
lateral view), the frame averaging per view (number of repeated
lateral views acquired and co-averaged at a particular rotation
step), and the number of rotation steps during a regular rota-
tion range (180 ) scan. Increasing any of these parameters
results—up to a certain limit—in reduced noise and/or
increased resolution in the final images, as well as a longer
scan time. The relative effect of varying these parameters
Fig. 2 (Modified from Ref. [9]) Sagittal (left) and transaxial (right) views of mouse
mandibles from deficient in dentin sialophosphoprotein (dspp / , a) and
heterozygote (b) animals, heterozygote animals show a normal teeth
development. The shorter secretional and maturation stages of enamel
development in dspp / are evidenced by the shorter span (indicated by
arrows) of the incisor from the cervical loop to the point enamel reaches its
final mineral density. Mandible volumes have been reoriented from their original
positions during scanning for an optimal sagittal view of the incisors (sagittal
plane marked by white line on transaxial view on b). An optimal orientation of the
same heterozygote jaw volume for a best molar view needs a separate
reorientation (c, note that the sagittal plane now goes through the middle of
the molars). This is because the mouse molars are located on an alveolar ridge
with an oblique orientation with respect to the incisors (evident in a 3D rendering
of the jaw, d)
(e.g., doubling the exposure time or the frames per view)

cannot be extrapolated from one system to another. For exam-
ple, certain manufacturers generally suggest acquisition using
one or two frames per view, while others suggest several, which
possibly reflects the geometry variability between systems.
7. 80 degree- vs. 360 degree-rotation range: Whereas acquisition
of lateral views over 180 degrees of total rotation provides the
full information needed for the 3D reconstruction of a
microCT (or any CT) volume, similar acquisition over 360
enables correction of some asymmetrical artifacts. These scans
can have significantly reduced noise levels, all other conditions
being equal, but also take twice as long and require a higher
computing power.
8. Contrast-enhanced specimens: A higher setting (70–90 Kvp,
depending on the thickness of the stained tissue) is commonly
used, to match the opacity of the contrast agent.
9. Imaging is most commonly performed in a vertical batch scan
fashion. Currently, some systems provide the option of using an
automatic sample loader, enabling consecutive batch scans of
several holders loaded onto a tray after pre-programming.
10. After all the conditions for the scan (or batch scan) have been
set, the parts of the holder to be scanned should be defined and
labelled. This is done based on a scout view (preliminary lateral
view of the holder). A directory for storage of images and
related specimen information is defined at this step.
11. Phantoms for mineral density calibration (most commonly
mineral-analogue rods of specific densities from a selected
range) should be separately imaged under identical conditions
(including holder and medium) to those used for the speci-
mens. The diameter of the mineral-analog rods should be
approximately the same as that of the scanned bone or dental
tissue. Soft tissue or liquid or mouting materials surrounding
the scanned sample should be matched approximately in the
phantom scan, since surrounding medium changes recon-
structed attenuation [10]. Some systems use built-in algo-
rithms for mineral density calibration for a specific beam
energy and voxel resolution scan, rather than the experimental
values obtained by the use of a phantom.
12. In vivo microCT: Reports of data from dental tissues from
in vivo systems are less numerous than reports from bone
[11–14]. The reason is the limitation in the maximum attain-
able resolution due to the big size of the imaged object (the
whole head of the anesthetized small animal), as well as the
limitation in radiation that can be delivered to the animal/time
it can stay anesthetized. The technical issues involved with this
mode are discussed in an excellent review by Badea et al. [1].
3 Methods
3.1 3D 1. During image acquisition, the scanner captures lateral projec-

Reconstruction of the tions of the object inside the holder, as they rotate along with
Scanned Volume the stage on which they are supported. The resulting file (raw
data) at the end is either a tiff, dicom or other format series of
images, or a proprietary format single file. Every system comes
with its own software for a 3D volume reconstruction (see Note
7) from the lateral projections (or the proprietary format single
file that contains them). Depending on the system, the user at
times needs to set values for some reconstruction parameters,
such as beam hardening and ring artifact correction, or selec-
tion of a range of attenuation coefficient values to employ as the
intensity window for grayscale values in the reconstructed vol-
ume. The output of the 3D reconstruction (the correct term is
reconstructed volume, rather than “scan”) is most of the time
also a series of images (tiff, bmp, jpeg) or again a single propri-
etary format file.
3.2 Post-processing 1. Reconstructed volumes are most commonly processed by the

of the Reconstructed manufacturer’s own densitometry and morphometry proces-
Volume sing software but can also be imported into a third-party soft-
ware after an appropriate format conversion. Caution has to be
exercised if a format conversion occurs, as the calibration of
grayscale values of images to mineral density may not be possi-
ble after it.
(a) Two commonly overlooked preliminary processing fea-
tures that can greatly facilitate and/or standardize the
analysis are (1) the possibility for digital reorientation of
volumes into a standard position before further processing
(Fig. 2) and (2) the possibility for digital resampling
(reduction of the image voxel resolution).
l Digital reorientation of volumes: it is very important to
set all the reconstructed volumes of specimens in a
standard orientation and work with those new files,
both for correctly understanding spatial relationships
between structures on 2D cross sections and for stan-
dardizing the ROIs for reproducible densitometry and
morphometry data extraction. This complements the
effort for a similar orientation of the specimens during
packing them for scanning, as described above. This
kind of software (commonly available as part of the 3D
morphometry software) works by (a) rotation of
images on the x, y, and z plane, (b) generation of new
reference planes based on coordinates of landmarks, or
(c) spatially manipulating the reference planes in a 3D
rendering of the volume, while the latter is kept in a

standard position.
l Digital resampling: is the process of averaging the
mineral density of neighboring voxels into a single
larger size voxel. It can make the analysis process effi-
cient, computer power-wise, for large (high resolu-
tion/big specimen size) volumes. For instance,
resampling to double the voxel size (known as a
2 2 binning of the volume—not to be confused
with binning of the detector elements) results in the
reduction of the imaging file size by a factor of 8. The
image resolution has been shown only to marginally
decrease compared to the original data upon digital
resampling.
(b) The most common kind of data in odontogenesis studies
of dental tissues and the jaw are visual, densitometric, and
volumetric ones. However, measurements included in
packages of 3D morphometry analysis meant primarily
for bone or materials can be performed successfully, if
they describe geometric properties similar to those in the
prototype application.
(c) Visual data: one can illustrate mineralization/develop-
mental changes between specimens in either a 2D (gray-
scale cross sections exported from viewing or processing
software) or 3D (3D rendering of the whole volume or a
part of it, after appropriate segmentation of tissues from
the background; see below) mode. In the 2D case, a similar
orientation of the specimen volumes is especially impor-
tant, as already noted. Care has to be exercised in the case
of 3D rendering of an examined volume, as structures on
the images can appear bigger or thicker/more mineralized
just by slightly varying the threshold selected for the
tissue-background segmentation. Most microCT proces-
sing software programs have a built-in 3D rendering func-
tion; otherwise exporting to a third-party software is
always possible. The tissue-background segmentation is
described below. Note that the 3D rendering of the recon-
structed volume is sometimes falsely called “3D recon-
struction.” The latter term only refers to generation of the
reconstructed volume from the initially acquired lateral
projections.
l Definition of ROI(s) is done through user-drawn con-
tours including the area of interest on several slices
throughout this area (e.g., every five slices) and inte-
gration through interpolation of the individual con-
tours. The latter can be drawn in a free handlike
(aided by software tools for refinement of a first “draft”

shape) manner or using regular geometrical shapes
(circles, ovals, and rectangles) positioned and adjusted
for size. Multiple, not communicating with each other,
or exclusion (areas circumscribed in them excluded
from processing) ROIs can be defined in some types
of software. The ROI can be saved as a separate and
possible to modify file that can be automatically applied
to the volume from which it originated. Some software
programs provide the ability to automatically define the
boundaries of a mineralized (or X-ray—opaque after
staining) object with a “ROI-warp” (on a slice or in 3D
mode)-type function, but this function has to be used
cautiously when the background/object interface does
not represent a sharp transition.
(d) A very common question among users is whether to use a
2D or 3D analysis to report quantitative results. 2D anal-
ysis is based on a slice-per-slice measurements of tissue
segmented from the background inside the area defined
by the ROI. 3D analysis is based on direct measurements
in space within the reconstructed volume, based on 3D
algorithms and using no geometric assumptions about
underlying structures [15]. The results reported most
often in odontogenesis studies are comparative volumetric
(e.g., volume of dentin, enamel or the whole crown/root
of molars, or rodents’ incisors formed in the examined
groups of animals), densitometric (e.g., respective average
values for crown dentin or enamel or root dentin), or—
less often—morphometry (e.g., average thickness of den-
tin and enamel) ones. For the first two kinds, 2D and 3D
analyses give almost identical results; therefore the 2D
analysis should be preferred due to the shorter time it
takes to be completed.
(e) Calibration of the mineral density values in the recon-
structed volume (conversion of grayscale values to g/cc)
has to be done using data from the (phantom) calibration
scan following the manufacturer’s instructions for the
particular software. Most commonly this is done by calcu-
lating a grayscale average from two or more mineral-
analogue rods imaged under identical to the specimens’
conditions, then entering the values into a calibration
dialogue of the processing software. This step is not
needed for systems that perform automatic calibration,
based on algorithms dependent on the scan beam energy
and voxel resolution—upon reconstruction of the imaged
volume.
Fig. 3 (a–c) Horizontal, transaxial, and sagittal views of a first molar in a 1-month-old mouse mandible (6 μm
voxel size, plane intersections noted by respective lines on views). (d) Histogram of mineral densities within a
region of interest of the first molar crown (shown in c). The interfaces of background (pulp), dentin, and
enamel peaks in the histogram are close to the baseline; therefore the measurements for dentin and enamel
can be easily done using respective cutoffs, such as the ones indicated by the black arrows. It is always
advisable to be extracting histograms of mineral densities from the regions of interest and use them as the
basis for segmentation. Note that although the ROI is shown on a sagittal view here, it is most commonly
drawn on transaxial views (b) as these involve the least amount of interfacing with the adjacent teeth
(f) A method should be chosen for tissue-background seg-

mentation by setting a respective mineral density thresh-
old. The most widely used method is setting a global
threshold (1 g/cc value is used to segment the tissue
from the background—or adjacent tissue—throughout
the analysis set) based on (1) a resulting binarized image
that best matches the grayscale image, (2) the point of
inflection between the tissue and the background/adja-
cent tissue peaks on a histogram of mineral densities inside
the ROI, or (3) the Otsu algorithm (Otsu 1979). The
dependence of morphometry and densitometry measure-
ments on thresholding has been reviewed in several stud-
ies [2, 16, 17]. It is our opinion that the second option
(using the point(s) of inflection—Fig. 3d) provides more
reproducible results that can also be better compared
between different reports. Alternatively, (and less fre-
quently used), the threshold can be a local adaptive one,
meaning defined by an algorithm built into the analysis
software and varying within the set of images. The former
kind of threshold is based on localized analysis of density
and aims at minimizing the partial volume effect and
thickness biasing, albeit it is not reproducible from one
software program to the other.
l Parameters reported in microCT imaging of odonto-
genesis studies: The most commonly reported results
are mean density and volume of (crown or root) dentin
and enamel, as well as volume of the crown or the root,

in molars. In already developed molars (or incisors past
the maturation stage of enamel), no separate ROIs for
the two tissues are needed because mature enamel
values are higher than mature dentin values and all
measurements can be based on an appropriate back-
ground/dentin/enamel thresholding (Fig. 3). Back-
ground can be density values from pulp space voxels
or the area around the crown—in case a crown or pulp
volume is not needed, the ROI doesn’t have to follow
the enamel boundary. Jawbone microarchitecture data
have been reported less commonly, and in these studies
the definition of the region of interest can vary signifi-
cantly. A summary of the bone morphometry para-
meters reported, as well as definitions of the region of
interest in past studies, can be found in the review by
Faot et al. [3], while the general guidelines for report-
ing bone morphometry data have been summarized in
Bouxsein et al. [15].
l It is important to keep in mind that absolute values
calculated for density (parameters described above)
should not be compared between studies. Although
few relevant data have been published [18], it is most
likely that relative densities (e.g., the percentage that
molar crown dentin differs between two different
mouse genotypes) should be comparable across sys-
tems under similar imaging conditions, but absolute
values (g/cc) may not be so. The same is not true for
volumetric data, which are expected to be reproducible
for same kind of specimens (with the exception of a
small variability due to possible segmentation differ-
ences) across studies.
(g) When reporting data, the system type and settings for the
scans (voxel size, filter, xray applied voltage, frames per
view, rotation step and rotation orbit – 180 or 360
degrees) have to be included. In a similar way, the type
of software used for reconstruction and processing has to
also be included, along with the reconstruction settings
(e.g., beam hardening or ring artifact correction factor),
the method for mineral density calibration, and the
method for definition of the ROIs and segmentation
used. Finally, if morphometry parameters were deter-
mined, the report has to include whether the measure-
ment was in a 2D or 3D mode and, ideally, the geometric
principle upon which these parameters are calculated by
the particular type of software (see also [15] guidelines for
bone microCT analysis).
4 Notes
1. The beam-hardening artifact in CT is due to the fact that

polychromatic radiation that traverses a dense or thick object
becomes progressively richer in high-energy photons, hence
more penetrating or “harder.” The artifact is expressed as a
cupping effect on the edge of medium density thick objects or
white/dark streaks originating from edge points of dense
objects [19].
2. A common mistaken assumption about microCT analysis is
that the mineral density calibration does not matter if no
mineral density but only volumetric results are to be reported.
Failure to standardize the holder/environment of the speci-
mens throughout a series of scans and perform a calibration for
mineral density scan of phantoms under the analysis conditions
is not a good practice. A lack of calibration between scans
makes the selection of a meaningful threshold for volumetric
measurements problematic.
3. The voxel size at a certain magnification depends on the geo-
metric magnification and the detector element size. The reso-
lution under the same conditions also depends on the
geometry of the system overall and the contrast that results
from the beam energy of a system. Even under optimized rest
of conditions (exposure time, rotation step, frame averaging),
the real resolution can be three or four times lower than the
voxel one, especially when the voxel size approaches the sys-
tem’s lower limit.
4. Binning mode of the detector: magnification of an object
during microCT imaging is defined by both the relative dis-
tances between source/object/detector and the detector bin-
ning mode. All modern systems have such different settings for
the latter, namely, an unbinned (1 1—correspond to a dis-
crete signal from every detector element) mode that offers the
maximum magnification and a 2 2 or even 4 4 binned
mode. Detector binning results in more efficient photon col-
lection and shorter scan times, but FOVs become, respectively,
half (for 2 2) or one fourth (for 4 4) that of the
unbinned mode.
5. It has to be noted that the region of signal-response (attenua-
tion) linearity for any microCT volume can be only calibrated
for a certain range of mineral densities [20], that is, measure-
ments cannot be quantitative for low and high densities within
the same volume. The linear region of attenuation and mineral
density (hence, the range of mineral density calibrated) has to
be chosen according to the density of the structures of interest
within the imaged volume.
6. Please note that the term “3D reconstruction” has been falsely
used to describe a 3D rendered image generated from the 3D
volume, not the generation of a 3D volume generated from
lateral projections after reconstruction as is correct. A 3D
volume is a file that contains coordinates for all the voxels
within the scanned area and a value for the X-ray absorption
coefficient (representing mineral density) associated with each
voxel.
References
1. Badea CT, Drangova M, Holdsworth DW, 10. Salmon PL, Liu X (2014) MicroCT bone den-
Johnson GA (2008) In vivo small-animal imag- sitometry: context sensitivity, beam hardening
ing using micro-CT and digital subtraction correction and the effect of surrounding media.
angiography. Phys Med Biol 53(19): The Open Access Journal of Science and Tech-
R319–R350 nology 2, Article ID 101142, 25 pages.
2. Kuhn JL, Goldstein SA, Feldkamp LA, Goulet https://doi.org/10.11131/2014/101142
RW, Jesion G (1990) Evaluation of a micro- 11. Zhao N, Liu Y, Kanzaki H, Liang W, Ni J, Lin J
computed tomography system to study trabe- (2012) Effects of local osteoprotegerin gene
cular bone structure. J Orthop Res 8 transfection on orthodontic root resorption
(6):833–842 during retention: an in vivo micro-CT analysis.
3. Faot F, Chatterjee M, de Camargos GV, Orthod Craniofac Res 15:10–20
Duyck J, Vandamme K (2015) Micro-CT anal- 12. Nan Ru, Sean Shih-Yao Liu, Li Zhuang, Song
ysis of the rodent jaw bone micro-architecture: Li, Yuxing Bai (2013) In vivo microcomputed
a systematic review. Bone Rep 2:14–24 tomography evaluation of rat alveolar bone and
4. Swain MV, Xue J (2009) State of the art of root resorption during orthodontic tooth
Micro-CT applications in dental research. Int movement. The Angle Orthodontist 83
J Oral Sci 1(4):177–188 (3):402–409
5. Parkinson CR, Sasov A (2008) High- 13. Furfaro F, Ang ESM, Lareu RR, Murray K,
resolution non-destructive 3D interrogation Goonewardene M (2014) A histological and
of dentin using X-ray nanotomography. Dent micro-CT investigation in to the effect of
Mater 24(6):773–777 NGF and EGF on the periodontal, alveolar
6. Zanette I, Enders B, Dierolf M et al (2015) bone, root and pulpal healing of replanted
Ptychographic X-ray nanotomography quanti- molars in a rat model - a pilot study. Progress
fies mineral distributions in human dentine. Sci in Orthodontics 15:2. https://doi.org/10.
Rep 5:9210 1186/2196-1042-15-2
7. Naveh GR, Brumfeld V, Shahar R, Weiner S 14. Soundia A, Hadaya D, Esfandi N, de Molon
(2013) Tooth periodontal ligament: direct 3D RS, Bezouglaia O, Dry SM, Pirih FQ, Aghaloo
microCT visualization of the collagen network T, Tetradis S (2016) Osteonecrosis of the jaws
and how the network changes when the tooth (ONJ) in mice after extraction of teeth with
is loaded. J Struct Biol 181(2):108–115 periradicular disease. Bone 90:133–141
8. Metscher BD (2009) MicroCT for comparative 15. Bouxsein ML, Boyd SK, Christiansen BA,
morphology: simple staining methods allow Guldberg RE, Jepsen KJ, Muller R (2010)
high-contrast 3D imaging of diverse Guidelines for assessment of bone microstruc-
non-mineralized animal tissues. BMC Physiol ture in rodents using micro-computed tomog-
9:11 raphy. J Bone Miner Res 25(7):1468–1486
9. Verdelis K, Szabo-Rogers HL, Xu Y et al 16. Ding M, Odgaard A, Hvid I (1999) Accuracy
(2016) Accelerated enamel mineralization in of cancellous bone volume fraction measured
Dspp mutant mice. Matrix Biol by micro-CT scanning. J Biomech 32
52-54:246–259 (3):323–326
17. Hara T, Tanck E, Homminga J, Huiskes R 19. Brooks RA, Di Chiro G (1976) Beam harden-
(2002) The influence of microcomputed ing in x-ray reconstructive tomography. Phys
tomography threshold variations on the assess- Med Biol 21(3):390–398
ment of structural and mechanical trabecular 20. Nuzzo S, Peyrin F, Cloetens P, Baruchel J,
bone properties. Bone 31(1):107–109 Boivin G (2002) Quantification of the degree
18. Verdelis K, Lukashova L, Atti E et al (2011) of mineralization of bone in three dimensions
MicroCT morphometry analysis of mouse can- using synchrotron radiation microtomography.
cellous bone: intra- and inter-system reproduc- Med Phys 29(11):2672–2681
ibility. Bone 49(3):580–587
Chapter 29
Transmission Electron Microscopy (TEM) and Scanning

Electron Microscopy (SEM) for the Examination of Dental
Hard Tissues
Victor E. Arana-Chavez and Leticia S. Castro-Filice
Abstract
This chapter describes laboratory protocols for TEM and SEM approaches allowing the examination of the
dental hard tissues’ constituents at the ultrastructural level. TEM has the highest resolution power to
examine the cellular and extracellular matrix ultrastructure inside a given sample, detecting the presence,
location, and quantification of organelles related to the metabolism of the cell type as well as membrane
specializations. SEM allows the observation of the sample surface, for examining dimensional topography
and distribution of exposed features.
Key words SEM, TEM, Microscopy protocols, Dental hard tissues
1 Introduction
Electron microscopy is a very important tool for the study of

biological samples. Indeed, the fine details of cells and extracellular
matrix are only possible to be discerned by applying high-power
examination with the employment of an electron beam instead of
light or laser. There are two main types of electron microscopy, the
transmission electron microscopy (TEM) and the scanning electron
microscopy (SEM). TEM has the highest resolution power to
examine inside a given sample the cellular and extracellular matrix
ultrastructure, detecting the presence, location, and quantification
of organelles related to the metabolism of the cell type, as well as
membrane specializations. SEM allows for the observation of the
sample surface, by examining dimensional topography and distri-
bution of exposed features; as SEM has high-resolution power and
large focus depth, images appear to look with a three-dimensional
view. Chemical analysis is also performed in conjunction with
electron microscopy. An example is the immunocytochemistry
technique called immunogold labeling, which uses antibodies
325
326 Victor E. Arana-Chavez and Leticia S. Castro-Filice
conjugated with gold particles. Since electron microscopy has a

compromise with ultrastructural preservation, immunocytochem-
istry is also considered a form of “biochemistry on section.”
Tooth development comprises a sequence of events in which
electron microscopy allows the view of some particular steps.
Indeed, most of the current knowledge on tooth development
derive from ultrastructural studies. As teeth and their periodontal
tissues possess a high mineral component, the tissue processing
must be carefully done in order to preserve all the cellular and
extracellular matrix components. This chapter presents laboratory
protocols for both TEM and SEM approaches, which are standard
in our laboratory, including key details that make the difference for
getting well-preserved tissues, thus allowing the examination of
their constituents at the ultrastructural level.
2 Materials
All solutions should be prepared by using ultrapure water (prepared

by purifying deionized water, to attain a sensitivity of 18 MΩ-cm at
25 C) and analytical grade reagents.
2.1 Stock Solutions (a) 0.1 M cacodylate buffer. Prepare 4.28 g trihydrate sodium
cacodylate (Na(CH3)2As023H2O) in 50 mL of ultrapure
water, and adjust to pH 7.4 with 0.2 M sodium hydroxide
(NaOH). Make up to 200 mL with ultrapure water. Store at
4 C.
(b) 20% formaldehyde. Dissolve 3 g of paraformaldehyde (CH2O)
in 15 mL of ultrapure water in an Erlenmeyer flask by heating
to 60–70 C on a stirring hot plate. Add drops of 0.1 N
sodium hydroxide and stir until the solution clears. This prep-
aration must be carried out under a fume hood equipped with
an exhaust system. Allow the solution to cool to room
temperature.
(c) 25% glutaraldehyde. Glutaraldehyde (electron microscopy
grade) is sold in sealed ampoules containing 10 mL or in
screw cap bottles with up to 100 mL in concentrations ranging
from 8% to 70%; we use 25% glutaraldehyde in our protocols
for preparing the primary fixatives.
(d) 4.13% EDTA. Dissolve 41.3 g ethylenediaminetetraacetic acid
(EDTA) disodium salt (Na2C10H14O8·2H2O) in 900 mL
ultrapure water. Bring to pH 7.2 by adding 4.3 g NaOH,
pellets. Make up to 1000 mL with ultrapure water. Store at
4 C.
(e) 1% osmium tetroxide (OsO4) in 0.1 M cacodylate buffer. Dis-
solve 1 g of crystalline osmium tetroxide in 100 mL of 0.1 M
TEM and SEM for Dental Hard Tissues 327
cacodylate buffer under a fume hood equipped with an

exhaust system. Then, aliquot into small bottles and store at
20 C.
2.2 Primary Fixatives A. For morphological ultrastructure: 2% glutaraldehyde and 2.5%

formaldehyde. Add 8 mL of 25% glutaraldehyde, followed by
12.5 mL of 20% formaldehyde, and then sufficient 79.5 mL of
0.1 M cacodylate buffer to make 100 mL of the primary fixative
solution. Adjust to pH 7.4 with drops of 0.2 N HCl.
B. For immunocytochemical labeling: 0.1% glutaraldehyde and
4% formaldehyde. Add 0.4 mL of 25% glutaraldehyde, followed by
20 mL of 20% formaldehyde, and then sufficient 79.6 mL of 0.1 M
cacodylate buffer to make 100 mL of the primary fixative solution.
Adjust to pH 7.4 with drops of 0.2 N HCl.
The fixative solutions must be prepared at the same day of
fixation. However, in some cases it can be aliquoted to small bot-
tles, and store at 20 C to maintain the stability for considerable
periods. On the day of fixation, the bottle must be thawed prior to
use, since diffusion of fixative into tissues is faster at room
temperature.
3 Methods
3.1 TEM (See Note 1) Although tissues are well fixed by perfusion, the most commonly
used method is immersion. A living specimen should be processed
3.1.1 Primary Fixation
as soon as collected and immersed in the first fixation (primary
fixative) bath where it is immediately cut out in small pieces
1–2 mm long. Specimens are dissected out at room temperature
and quickly immersed in the fixative according to the type of
ultrastructural study, i.e., morphological or immunocytochemical.
For morphological studies, the best fixative is 2% glutaralde-
hyde + 2.5% formaldehyde (freshly prepared from paraformalde-
hyde) buffered at pH 7.4 with 0.1 M sodium cacodylate. For
immunocytochemical studies, the fixative must be 0.5% glutaralde-
hyde + 4% formaldehyde (freshly prepared from paraformaldehyde)
(for immunocytochemical studies) buffered at pH 7.2 with 0.1 M
sodium cacodylate. Fixation time is variable, depending on the size
and the density of tissue. For hard tissues, the specimen is kept into
the fixative by using a tissue rotator at a low speed for 6 h at room
temperature and then left at 4 C (refrigerator) overnight.
Although the fixation protocol above is standard in most
laboratories, we have established a protocol by using microwave
irradiation in order to improve the fixative diffusion into tissues
[1–4]. With this procedure, we always use a fixative with a low
concentration of glutaraldehyde (0.1%). As the quality of preserva-
tion after microwave fixation is higher than after the conventional
method, the fixed tissues can be used for both morphological and
immunocytochemical studies. The protocol for microwave fixation

is as follows:
1. Specimens are dissected out and quickly immersed in 0.1%
glutaraldehyde + 4% formaldehyde (freshly prepared from para-
formaldehyde) buffered at pH 7.2 with 0.1 M sodium cacody-
late at room temperature.
2. All the soft tissues covering the mineralized tissues should be
gently removed, and some regions of the overlying bone
should be partially broken by using bone-cutting forceps or
shears at regions far from the chosen for ultrastructural ana-
lyses. In rat mandibles, for example, the tip of incisor is broken
for exposing the dental pulp and opening a way for fixative
penetration. Whether the region of interest is the molar region,
the mandibular ramus must be broken with the bone-cutting
forceps; some thin layers of the bone can be broken with a
dental excavator.
3. Specimens are immersed in a beaker containing 40 mL of the
fixative solution at room temperature, which is subsequently
placed in a large (20 20 cm) glass recipient full of ice and
immediately taken inside a Pelco 3440 laboratory microwave
oven (Ted Pella, Redding, CA, USA) (see Note 2).
4. The temperature probe of the oven is submersed into the
fixative solution (it is usually at 18–20 C –room temperature).
5. The specimens are immediately exposed to microwave irradia-
tion at a 100% setting for three periods of 5 min; the tempera-
ture is programmed to a maximum of 37 C (at 100% setting,
microwave irradiation is continuous in the Pelco microwave
oven) (see Note 3).
6. Between the three periods of microwave irradiation, specimens
are cleaned, broken, and removed from the surrounding areas
and placed into a fresh fixative solution.
7. After fixation with microwaves, specimens are transferred into
fresh fixative and left in it in a tissue rotator at a low speed for
4 h and then overnight at 4 C.
3.1.2 Washing Samples are extensively washed in 0.1 M cacodylate buffer in order
to remove the excess of fixative. 4, 20 min each time.
3.1.3 Decalcification Decalcification of mineralized specimens carried out in an aqueous

solution of 4.13% EDTA [5] under microwave (MW) irradiation
for a cumulative time of 15 h in a Pelco 3440 laboratory MW oven.
Specimens are placed in a beaker containing 25 mL of the decalcify-
ing solution, which are placed in a larger glass recipient filled with
crushed ice. The temperature probe is submersed in the EDTA
solution, and the specimens are immediately exposed to MW
irradiation at 100% setting for periods of 15 min with the tempera-

ture programmed to a maximum of 37 C. The decalcifying solu-
tion is changed every hour and the ice replaced as required.
3.1.4 Trimming and After decalcification, specimens are divided into segments accord-
Washing ing to the orientation for inclusion and further sectioning and then
washed extensively in 0.1 M sodium cacodylate buffer, pH 7.2.
3.1.5 Post-Fixation Specimens are left into 1% aqueous OsO4 in 0.1 M cacodylate
buffer for 2 h at 4 C (not for immunocytochemistry).
3.1.6 Dehydration The dehydration process of a biological sample needs to be done

very carefully. Initially, the sample is immersed in 30% ethanol and
then bathed in solutions of increasing ethanol concentrations to
pure ethanol. The dehydrating times should be adjusted to size and
kind of tissue.
30% ethanol—1 5 min.
3.1.7 Embedding Once the sample is dehydrated, the infiltration process must be
initiated according to the type of resin epoxy (Spurr) for morpho-
logical ultrastructural studies or acrylic (LR White) for immunocy-
tochemical post-embedding labeling.
Spurr Resin Whether the tissues are embedded in the epoxy resin Spurr, pure
acetone must be used after 100% ethanol (pure acetone 4
10 min). And then:
1:1 Acetone/Spurr resin 1:30 h.
1:3 Acetone/Spurr resin 3 h.
Pure Spurr resin overnight.
Next morning, change out to new pure resin for 1–3 h.
Polyethylene capsules or silicone rubber embedding molds are
placed in a holder, and numbered strips of paper are inserted. A
drop of fresh resin is placed in the capsules, and the specimen is
transferred to the appropriate capsule or molds.
The “blocks” are cured for 48 h in a 60 oven.
LR White (Hard Grade) Whether an acrylic resin LR White is used (for immunocytochemi-
Resin cal studies):
100% ethanol 1:1 LR White resin (overnight at 4 C).
100% ethanol 1:3 LR White resin (3 h at 4 C).
LR White (2 h at 4 C).
LR White (overnight at 4 C).
LR White (1 h at room temperature).
Embedding is carried out in gelatin capsules in which the
bottom must be previously flattened with the use of a hot plate
for better orientation of specimens. After three to five drops of LR
White resin, place the specimen at the bottom, then fill the gelatin
capsule with resin, and close it well to seal the capsule to avoid the
oxygen which may interfere with polymerization. Then, leave to
polymerize at 60 C for 48–72 h.
3.1.8 Sectioning The block is cut into semithin sections (1–2 μm) with a glass knife,
using an ultramicrotome. The thick sections are stained with tolui-
dine blue and then examined with a light microscope for selecting
regions, which are trimmed for ultrathin sectioning. Ultrathin sec-
tions (50–80 nm) are made using a diamond knife and collected on
metal-mesh “grids” (see Note 4).
3.1.9 Post Staining l Stain grids in uranyl acetate for 2 h.

l Stain grids in lead citrate for 5 min.
Formvar film is useful for the support of ultrathin sections on the
finer mesh grids (see Note 5).
3.2 SEM Fix tissue in primary fixative as usual.

3.2.1 Fixation
3.2.2 Washing Wash tissue in 0.1 M cacodylate buffer.
3.2.3 Dehydration Dehydrate with a graded ethanol series by subsequent exchanges of

the following dilutions in distilled water as follows:
30% ethanol, 1 5 min.
3.2.4 Desiccation Although critical point drying (CPD) is the most common and
universal method for desiccation of dehydrated biological speci-
mens, it may introduce some artifacts, besides that it needs specific
equipment. Then, it is possible to use chemicals as a good and
efficient alternative. We largely use the chemical dehydrant hexam-
ethyldisilazane (HMDS) for desiccating hard tissues with compara-
ble results those of CPD.
After the last bath in 100% ethanol, tissues are quickly
immersed in pure HMDS for 15 min under a fume hood equipped
with an exhaust system. Then, leave the sample on a filter paper,
cover it with a glass plate, and leave under the hood until the
HMDS evaporates off.
3.2.5 Sputter Coating After desiccation, specimens are coated with a thin layer of a con-
ductive metal that minimizes damage to specimens against the
electron beam and improves the topographical contrast when sec-
ondary electron detection is used. Specimens will be mounted onto
aluminum stubs and sputter coated in a vacuum with an electrically
conductive 25-nm-thick layer of with gold/palladium or store
desiccated until coated.
4 Notes
1. Modifications to the basic TEM protocol can be applied to

achieve many different goals. Each and every sample that
comes to us is evaluated, and the procedure is adjusted to
make sure the best results will be obtained.
2. The fixation can be done in a regular microwave oven. How-
ever, it should be previously tested with 50 mL of distilled
water for knowing the time in which this amount of fixative
will reach 37 C in the given oven.
3. After knowing the time after which the fixative solution reaches
37 C, the irradiation times should be always the same. Do
several irradiations to complete the total time of 15 min
(change the fixative solution between each irradiation).
4. For undecalcified specimens some grids containing mineralized
specimens could be decalcified in 4.13% EDTA solution for
30 min and rinsed in distilled water for 10 min.
5. Using of support film is ideal for those applications requiring
large viewing areas without grid bar interference.
References
1. Jimenez-Pellegrin C, Arana-Chavez VE (2007 microscopy analysis combined with immunola-

Aug) Root resorption repair in mandibular first beling of osteopontin. Am J Orthod Dentofac
premolars after rotation. A transmission electron Orthop 132(2):230–236
2. Moretto SG, Azambuja N Jr, Arana-Chavez VE, 4. Laboux O, Dion N, Arana-Chavez V, Ste-Marie
Reis AF, Giannini M, Eduardo Cde P, De Freitas LG, Nanci A (2004) Microwave irradiation of
PM (2011) Effects of ultramorphological changes ethanol-fixed bone improves preservation,
on adhesion to lased dentin-scanning electron reduces processing time, and allows both light
microscopy and transmission electron microscopy and electron microscopy on the same sample. J
analysis. Microsc Res Tech 74(8):720–726 Histochem Cytochem 52(10):1267–1275
3. Massa LF, Ramachandran A, George A, Arana- 5. Warshawsky H, Moore G (1967) A technique
Chavez VE (2005) Developmental appearance for the fixation and decalcification of rat incisors
of dentin matrix protein 1 during the early den- for electron microscopy. J Histochem Cytochem
tinogenesis in rat molars as identified by high- 15(9):542–549
resolution immunocytochemistry. Histochem
Cell Biol 124(3–4):197–205
Part V
Protocols to Study Dental Diseases

Chapter 30
Rodent Dental Fluorosis Model: Extraction of Enamel

Organ from Rat Incisors
M. Suzuki and J. D. Bartlett
Abstract
Chronic fluoride overexposure can cause dental fluorosis. Dental fluorosis is characterized by porous and
soft enamel that is vulnerable to erosion and decay. Animal models often contribute to clinical applications
by addressing pathogenic questions of disease. To study dental fluorosis, rodent models have been
employed because rodent incisors erupt continuously and every stage of enamel development is present
along the length of the rodent incisor. Here we present a protocol to induce dental fluorosis in mouse and
rat and describe the procedure for extraction of stage specific enamel organ from rat mandibular incisors.
Key words Dental fluorosis, Fluoride, Amelogenesis, Enamel organ, Animal model, Rat, Mouse
1 Introduction
Fluoride is a specific and effective caries prophylactic. To prevent

dental caries, addition of fluoride to drinking water at a concentra-
tion of 0.7 ppm is recommended by the US Public Health Service
(PHS) [1]. However, chronic fluoride overexposure during preer-
uptive enamel development can cause dental fluorosis. Severe den-
tal fluorosis is characterized by porous and soft enamel that is
vulnerable to erosion and decay [2]. Since rodent incisors erupt
continuously and every enamel development stage including the
secretary stage and maturation stage is present along the length of
the incisor, the rodent model is a valuable tool to study enamel
development in mammals. Incisors are categorized as mandibular
or maxillary. Their respective enamel organs can be segregated into
the secretory stage or maturation stage. To induce dental fluorosis
in rat or mouse, animals are provided with water containing fluo-
ride as sodium fluoride (0–100 ppm) ad libitum. After 6 weeks,
animals are euthanized, and incisors are dissected and subjected to
experiments such as quantitative fluorescence (QF) assay to evalu-
ate fluorosis severity [3], microhardness measurements [4], RNA
extraction for qPCR [5], and immunohistochemical analysis [6].
335
336 M. Suzuki and J. D. Bartlett
Here we describe (1) the protocol to induce dental fluorosis in

rodents and (2) the incisor dissection procedure and stage-specific
extraction of enamel organ from secretory stage and maturation
stage.
2 Materials
2.1 Animals Mouse: C57BL/6 mice (6-week-old) (Charles River Laboratories,

Wilmington, MA).
Rat: Sprague-Dawley rats (6-week-old) (Charles River
Laboratories).
2.2 Fluoride- Rodent diet. 1/200 Pellets. Product No. F1515, AIN-76A (Bio-Sev,
Free Food Frenchtown, NJ).
Store diet at 4 C (Bring to room temperature before feeding).
2.3 Fluoride Water Autoclaved distillated water.

Sodium fluoride (NaF), Cat. No. S299-100 (Fisher Scientific).
Disposable Vacuum Filter/Storage System, 0.22 μm, 1000 mL
(Corning).
1. To prepare 50 ppm of fluoride water, add 0.111 g of NaF to 1 L
of water.
2. To prepare 100 ppm of fluoride water, add 0.222 g of NaF to
1 L of water.
3. Filter fluoride water using 0.22 μm disposable Vacuum Filter/
Storage System.
4. Fluoride water can be stored at 4 C for 2–3 days.
5. Bring fluoride water to room temperature before feeding.
2.4 Euthanasia Carbon dioxide (CO2).

Euthanasia chamber.
2.5 Dissection of Dissecting microscope.

Incisors 1.5 mL microcentrifuge tubes (two tubes, one for secretory (SEC)
stage and one for maturation (MAT) stage per animal).
Ice.
Flat frozen ice pack.
Rodent small animal guillotine.
Anatomy Dissection Kit (including general surgical scissors, for-
ceps, scalpel blade).
Bone-cutting shears.
Petri dish (10 cm).
Ice-cold sterile PBS.
Rat Enamel Organ Extraction 337
Cotton gauze.
Paper towels.
70% Ethanol (for cleaning instruments).
500 μL of TRIzol in one 1.5 mL microcentrifuge tube per sample
(for RNA extraction from mandibular incisor).
Fixation solution (4% paraformaldehyde in PBS).
Decalcification solution (20% sodium citrate and 10% formic acid in
distillated water).
3 Methods
3.1 Induction of All animals are treated humanely, and all handling procedures
Dental Fluorosis in should be approved by the Institutional Animal Care and Use
Rodents Committee (IACUC) at the institute where experiments are carried
out.
1. Six-week-old rodents (mice or rats) are divided into three
groups (Fluoride; 0, 50, 100 ppm).
2. Each group is provided with water containing 0, 50, or
100 ppm fluoride as NaF ad libitum for 6 weeks.
Water is changed two to three times per week (see Note 1).
3. Animals are fed with fluoride-free chow (Bio-Sev).
4. After fluoride treatment for 6 weeks, animals are euthanized
with CO2 (see Note 2).
Figure 1 shows incisor phenotypes from rat (Fig. 1a) and
mouse (Fig. 1b) after 6 weeks fluoride drinking water treatment.
3.2 Dissection of 1. After CO2 euthanasia, remove the head from the body using a
Incisors small animal guillotine (see Note 3).
3.2.1 Removal of the 2. Remove the skin from the head by the use of surgical scissors.
Head from the Body
3.2.2 Removal of 1. Remove hemi-mandibles from the head using surgical scissors.
Mandible from the Head Be very careful to not break or cut mandibles near apical ends of
incisors.
2. Place mandibles in a petri dish on ice (or on an ice pack).
Keep them on ice and proceed to Subheading 3.3.
3.2.3 Removal of 1. Slice between maxillary incisors using a razor blade.

Maxillary Incisors from Place razor blade between maxillary incisors and guardedly
the Head slide razor through the palate to get beyond the apical end of
incisor.
Fig. 1 Incisor phenotype of rats and mice treated with fluoride. Rats (a) and mice (b) were provided with water
ad libitum containing 0, 50, or 100 ppm fluoride for 6 weeks. Pictures are of three animal incisors for each
group. In both rats (a) and mice (b), compared to the 0 ppm fluoride group, the tooth color was changed to
chalky white opaque in both 50 ppm and 100 ppm fluoride groups. In mice (b), attrition (indicated by arrow)
was observed in 50 ppm and 100 ppm groups, and white spots (indicated by arrow head) were detected in the
100 ppm group
2. Hold the snout on the lateral side (to hold the razor blade in
place), and swing the razor out to the side to snap out the
maxilla.
3. Wiggle maxillary bone fragment with incisor to remove from
surrounding tissue, and place in a petri dish on ice (or on an ice
pack). Wash away any blood using ice-cold PBS. Repeat for
other incisor.
4. Put hemi-maxillary bone with incisor into 15 mL tube contain-
ing 4% paraformaldehyde in PBS.
5. Proceed to fixation with 4% paraformaldehyde overnight at
room temperature.
6. Wash with PBS, and place in decalcification solution (20%
sodium citrate and 10% formic acid) at 4 C for 2 weeks fol-
lowed by embedding into paraffin.
3.3 Extraction of 1. Use scissors and gauze to remove as much tissue as possible
Enamel Organ (EO) from hemi-mandible.
from Mandibular 2. Wash with cold PBS to remove blood, and place the hemi-
Incisor (Rat) (See mandible in a petri dish on an ice pack.
Note 4) 3. Do this for both hemi-mandibles.
Rat Enamel Organ Extraction 339
Fig. 2 Schematic diagram of rodent hemi-mandible. The solid black reference

line between first (M1) and second molar (M2) demarcates the secretory (SEC)
and maturation (MAT) stages. The dashed line outlines the approximate position
of the incisor within the bone
4. Use bone shears (“pinch and twist” technique), scissors, and

forceps to tease the bone away from incisor labial surface (see
Note 5).
5. Once the labial surface is free, keep specimen in petri dish on an
ice pack.
The following procedure requires a stereomicroscope.
6. Observe the surface color and texture of enamel organ under a
stereomicroscope (see Note 6).
7. Use scalpel to test enamel. Only secretory stage (SEC) enamel
can be scratched, but maturation stage (MAT) is very hard and
cannot be scratched (the scalpel does not sink into MAT stage
enamel).
8. Use scalpel to mark between areas that can and cannot be
scratched. This divides SEC and MAT, which is between the
first and second molar (Fig. 2).
9. Scrap SEC EO off enamel surface, and place into 1.5 mL
microcentrifuge tube containing TRIzol. Keep on ice.
10. Wipe scalpel with gauze/EtOH to minimize SEC/MAT
contamination.
11. Scrap MAT EO off enamel surface, and place into 1.5 mL
microcentrifuge tube containing TRIzol. Keep on ice.
12. Proceed to RNA extraction (see Note 7).
4 Notes
1. Depends on the amount of consumption by rodents.

2. Follow the IACUC guidelines.
3. For mice, surgical scissors also work.
4. Because of its small size, it is difficult to extract SEC and MAT
stages from mouse EO separately.
5. Be very careful when pulling bone off the tooth as EO may stick
to bone surface (red “goop” on tooth surface).
6. Secretory stage (SEC): translucent, grayish, and softer enamel.
Maturation stage (MAT): white, opaque, very hard, will not
mark; orange top coating.
7. Or store at 80 C ASAP so as not to degrade the RNA.
Acknowledgment
This work was supported by the National Institute of Dental and

Craniofacial Research of the National Institutes of Health under
award number R01DE018106 (JDB).
References
1. U.S. Department of Health and Human Services 4. Sharma R, Tye CE, Arun A, MacDonald D,
Federal Panel on Community Water Fluoridation Chatterjee A, Abrazinski T, Everett ET, Whit-
(2015) U.S. Public Health Service recommenda- ford GM, Bartlett JD (2011) Assessment of den-
tion for fluoride concentration in drinking water tal fluorosis in Mmp20 +/ mice. J Dent Res 90
for the prevention of dental caries. Public Health (6):788–792. https://doi.org/10.1177/
Rep 130(4):318–331. https://doi.org/10. 0022034511398868
1177/003335491513000408 5. Suzuki M, Bartlett JD (2014) Sirtuin1 and
2. DenBesten PK (1999) Biological mechanisms of autophagy protect cells from fluoride-induced
dental fluorosis relevant to the use of fluoride cell stress. Biochim Biophys Acta 1842
supplements. Community Dent Oral Epidemiol (2):245–255. https://doi.org/10.1016/j.
27(1):41–47 bbadis.2013.11.023
3. Everett ET, McHenry MA, Reynolds N, 6. Suzuki M, Bandoski C, Bartlett JD (2015) Fluo-
Eggertsson H, Sullivan J, Kantmann C, ride induces oxidative damage and SIRT1/
Martinez-Mier EA, Warrick JM, Stookey GK autophagy through ROS-mediated JNK signal-
(2002) Dental fluorosis: variability among dif- ing. Free Radic Biol Med 89:369–378. https://
ferent inbred mouse strains. J Dent Res 81 doi.org/10.1016/j.freeradbiomed.2015.08.
(11):794–798 015
Chapter 31
Three-Dimensional Assessment of Crown Size

and Eruption Space for Developing Third Molars:
Data Collection Techniques Based on Cone-Beam
Computed Tomography (CBCT)
D. F. Marchiori, G. V. Packota, and J. C. Boughner
Abstract
Third molar development and eruption are two related areas of major interest in dental research into the
etiology of “wisdom tooth” impaction. Third molars are not only an excellent model for studying dental
development but also of fundamental clinical importance because they are very frequently impacted.
Because the third molar is located in the distal-most region of the oral cavity, clinical access is relatively
challenging. With the increasingly widespread use of cone-beam computed tomography (CBCT) in
dentistry, studies and measurements of the third molar and its eruption area have become considerably
easier to do. Here we present a novel CBCT-based measurement methodology we developed for our recent
investigations that we hope will also be useful for the broader dental research community.
Key words Third molar, 3D imaging, Measurement techniques, Tooth development, Research
methods, Wisdom tooth impaction
1 Introduction
The third molar (M3) is an optimal model for studying dental

development because this distal-most molar is the only human
tooth that can be observed radiographically from initiation to
root completion [1]. Because the M3 is located in the distal and
less accessible regions of the oral cavity, its actual spatial position
relative to adjacent structures—such as the first (M1) and/or sec-
ond (M2) permanent molars—is not always accurately depicted by
conventional radiographic methods [2–4]. Simultaneously, the
technical limitations [5] of these conventional radiographic meth-
ods (e.g., elongation, distortion, superposition) traditionally used
in dental research also pose challenges for designing measurement-
based studies of M3 mineralization and eruption [6]. As a result,
3D imaging methods such as cone-beam computed tomography
341
342 D. F. Marchiori et al.
(CBCT) have begun to be adopted [7], or recommended [8], by

more recent studies assessing M3 development. CBCT imaging
allows not only a highly accurate analysis of M3 developmental
stages in any anatomical plane (e.g., axial, coronal, sagittal) but
also permits the M3 crown size and surrounding jaw space condi-
tions to be measured [2–4, 7], thus enabling new and valuable
insights into the factors that increase risk of M3 impaction.
Because CBCT has so much potential to allow considerable
advances in the areas of M3 development, eruption, and impaction,
we anticipate that CBCT will become a standard tool for dental
research [7]. With established measurement methodologies being
scarce in the literature, investigators aiming to study and measure
the M3 and its region are faced with the challenge of creating their
own metrics. The disadvantage of this variability among data col-
lection methodologies is that it is not only time-consuming but,
more to the point, difficult to compare among and build upon
reported results. This chapter aims to deliver technical guidelines
toward (1) standardizing CBCT-based measurement techniques
for permanent molars and their region of the jaw and (2) maximiz-
ing data consistency among studies. Here we present a study
method to assess and measure the mesiodistal and buccolingual
diameters of molar crowns as well as the length of the M3 eruption
space in the dental arch.
2 Materials
At the crux of this method is the ability to read digital format

computed tomographic (CT) images and thus generate 3D multi-
planar (e.g., coronal, sagittal, axial plane) (Fig. 1) views of an
individual’s anatomical structures (i.e., the M3 and its eruption
area). Such images permit linear measurements to be taken with
higher accuracy and precision compared to conventional 2D dental
radiographs [4, 9]. We used the proprietary software package Xoran
i-CAT (Imaging Sciences International, Philadelphia, USA), ver-
sion 3.1.62. This software is not unique, sharing common basic
tools and capabilities with other software of its kind (e.g., InVivo-
Dental, Anatomage) that allow CBCT scan files to be visualized and
measurements to be taken “in situ”, that is, without having to
export images to be studied or measured by other specialized
software. Each CT reader may have distinct computer hardware
and operating system requirements, although most appear to work
on some version of Windows. Thus your software of choice may
depend on your operating system or vice versa.
Despite the volumetric capabilities of CBCT images, for the
purposes of data standardization and thus reproducibility, our
method for measuring the M3 and its eruption area was developed
by studying single-image “slices” within a given plane (e.g., axial,
CBCT-Based Methods for Studying Third Molars 343
Fig. 1 The position of each anatomical plane (e.g., coronal, sagittal, axial) (a) generates corresponding 2D
non-volumetric images (b) which are then used for studying third molars. Red line, axial plane; yellow line,
sagittal plane; green line, coronal plane
sagittal, coronal). Therefore, all observations and measurements are

taken on the same plane (e.g., axial) while still using the sagittal and
coronal planes for spatial referencing. Measurements on volumetric
3D surfaces themselves may be possible, but such process is more
complex and may require not only several additional methodologi-
cal steps, but also sophisticated software (e.g., Amira, Avizo) to
create a volumetric (3D) rendering of the anatomical structures
under investigation.
To record descriptive observations and measurement data, we
suggest using software such as Microsoft Word or Excel to serve as
database, especially if you choose a CBCT software reader that does
not have the capability to save your notes and/or measurements.
Regularly backing up of your data to a separate and stable drive
(e.g., external hard drive) is also recommended.
3 Methods
3.1 Standard Initial Before proceeding to the measurement method itself, the 3D ori-
Configurations for entation of the individual’s head and jaws need to be adjusted.
Measuring the Third Using your preferred CBCT viewing software (refer to Materials),
Molar and Its Area perform the following initial steps:
1. Open the CBCT imaging file of choice. Make sure the indivi-
dual’s image has no technical imperfections (e.g., blurring,
overshining metallic objects) that may affect the proper obser-
vation of the structures under investigation (see Note 1).
2. Adjust the thickness of the image slices. This generally needs to
be done for each anatomical plane window: sagittal, axial, and
Fig. 2 First steps before measuring the dimensions of the third molar crown and its putative eruption area.
(A) Axial view window: move the coronal (green line) and sagittal (yellow line) planes to intersect each other
right on the center of the M1’s crow. Avoid deviations such as the one exemplified in (E). (B) Sagittal view
window: Initiate by placing the axial plane (red line) on the level of the occlusal surface of the M1 and M2 in the
oral quadrant under investigation. Next, move the axial plane downward to the level of the mesial and distal
interproximal contact points of the M1. (C) Coronal view window: make sure that the axial plane (red line) is
crossing the right and left M1s at their same superior-inferior anatomical level. Avoid tilting the subject’s head,
as exemplified in (D)
coronal. Thin slices allow a more accurate analysis of targeted

anatomical structures. For our CBCT-based method, slices
1 mm or thinner generate sufficiently detailed images. For
consistency, use the same slice thickness for all individual
images studied (see Note 2).
3. In the axial view window, move the coronal and sagittal planes
to intersect each other right on the center of the M1’s crown
(Fig. 2A) (see Note 3).
4. In the sagittal view window, the axial plane can be moved
superiorly or inferiorly. Move the axial plane to the occlusal
level of the M1 and M2 in the oral quadrant under investiga-
tion. A straight axial plane connecting the tips of the M1 and
M2 cusps is an important baseline to establish because it coin-
cides with the occlusal plane. We will herein refer to this
position simply as the occlusal level (OL) (Fig. 2B). From the
OL, you may subsequently move the axial plane upwards or
downwards as necessary, always remaining parallel to the OL
(see Note 4).
5. Next, still in the sagittal view window, move the axial plane
downwards to a level which coincides with the mesial and distal
interproximal contact points of the M1 and M2 (Fig. 2B). We
will herein refer to this position simply as the contact level

(CL) (see Note 5).
6. In the coronal view window, make sure that the axial plane is
crossing the right and left M1s at their same superior-inferior
anatomical level, that is, in the superior-inferior mid-third of
their crowns (which is the level where the interproximal contact
points are located), as determined in step 4 (Fig. 2C).
Once step 6 is done, the image is ready for the measurement
techniques that we describe next. This initial configuration step
needs to be done for all individuals’ images to create standard
conditions for a consistent and reliable reproducible data collection
method.
3.2 Third Molar For measuring molar crowns, some key spatial and anatomical
Crown Dimensions features need to be taken into consideration. In terms of molar
crown anatomy, for instance, it should be noted that the maximum
mesiodistal crown diameter (MDMAX) is normally found at a level
slightly superior to that level where the maximum buccolingual
crown diameter (BLMAX) is normally found (Fig. 3). For accurate
measurements, these anatomical features of the molar crown need
to be taken into account. For instance, M3s with their long axes
severely inclined (e.g., angulated) relative to the axes of the adjacent
M1 and M2 may be reasonably more difficult to measure (Fig. 3D).
The measurement method presented here will show, however,
alternative techniques for properly measuring such angulated
molars.
This measurement method is designed to assess molar crown
diameter in two dimensional aspects: mesiodistal (Fig. 4D) and
buccolingual (Fig. 4E). Assuming that steps 1–6 from section 3.1
above are completed, then follow the instructions below to mea-
sure the dimensions of the M3 crown.
3.2.1 Measuring the 1. First, determine whether or not the M3’s long axis is tilted in
Maximum Mesiodistal the mesiodistal aspect (see Note 6). If the M3’s long axis is
Crown Diameter (MDMAX) noticeably tilted (more common), follow step 2 before going
to step 3. If the M3’s long axis is not noticeably tilted (uncom-
mon), then skip step 2.
2. For M3s (or any other molars) that are noticeably tilted in the
mesiodistal aspect (Fig. 3D), a compensatory inclination of the
axial plane is necessary to see and accurately measure their
MDMAX in the axial view window. Perform the compensatory
inclination in the sagittal window view, as instructed in Fig. 5.
3. In the axial viewing window, identify the “anatomical”
mesial-most (MMan) point of the tooth’s crown (for details
see Fig. 3A, B) (see Note 8).
4. Next, identify the “anatomical” distal-most (DMan) point of
the tooth’s crown (for details see Fig. 3A, B) (see Note 8).
Fig. 3 Key anatomical features to be considered when measuring molars. (A) Illustrates the mesial-most (MM),
distal-most (DM), buccal-most (BM), and lingual-most (LM) points of a molar crown. (B) Note that when the
tooth is giroverted (e.g., rotated), the “anatomical” points (e.g., MMan, DMan, BMan, LMan) may not coincide
with the position of the “absolute” points (e.g., MMab, DMab, BMab, LMab). This occurs because, with rotation
of the crown, its “anatomical” surfaces (and points) will rotate to a new position. Since a rotated tooth is only
changed in terms of its position (not its size), its maximum mesiodistal (MDMAX) crown diameter is always
measured between its MMan and its DMan (blue arrows). Alike, its maximum buccolingual (BLMAX) diameter
must be always measured between its BMan and its LMan. (C) Note that the level used for measuring the
maximum mesiodistal crown diameter (MDMAX) differs from the level used for measuring the maximum
buccolingual crown diameter (BLMAX). (D) Illustrates how challenging crown measurements may be if the
molar under investigation is severely tilted
5. Use your software measurement tools to measure the M3 crown

from its mesial-most (MMan) point to its distal-most (DMan)
point (Fig. 3A, B). Provided that any required compensatory
inclination of the axial plane is made, this method may be used
to measure the M1 and M2 crowns as well (see Note 9).
3.2.2 Measuring the 1. First, determine if the M3’s long axis is tilted in the buccolin-
Maximum Buccolingual gual axis (see Note 6). If the M3’s long axis is noticeably tilted,
Crown Diameter (BLMAX) follow step 2 before going to step 3. If the M3’s long axis is
not noticeably tilted, then skip step 2.
Fig. 4 Measuring molar crowns on CBCT images. A, B, and C shows the permanent molars of a patient through
three distinct views: axial, sagittal, and coronal, respectivelly. Both the maximum mesiodistal crown diameter
(D) and the maximum buccolingual crown diameter (E) may be measured. For reproducibility of the method,
measurements are always done in the axial view window of the patient image. Note that this method may be
used to measure not only the M3 crown but the M1 and M2 crowns as well
Fig. 5 Compensatory inclinations of the axial plane. If the M3 is noticeably tilted in the mesiodistal aspect (A),
the axial plane (red line) at the level of the interproximal contact points of the M1 and M2 (contact level, or CL)
may not generate in the axial view window a proper image of the real dimensions of the M3’s crown (B, arrow).
Note that the M3’s crown is not seen in Fig. B above, only its root. In such cases, a compensatory inclination of
the axial plane is necessary (A, *) for allowing the maximum mesiodistal diameter of the M3’s crown to be seen
and measured in the axial view window (C, arrow). The circumference of the crown is now properly visible. The
ideal amount of inclination of the axial plane is reached when this plane is perpendicular to the long axis of the
M3. Following this compensatory inclination, move the axial plane to the crown’s contact level. Note that
compensatory inclinations of the axial plane may be also done for tooth tilted in the buccolingual aspect. In such
cases, use the coronal view window for promoting the compensatory inclination
2. For M3s (or any other molars) that are noticeably tilted in the
buccolingual aspect (Fig. 3D), a compensatory inclination of
the axial plane is necessary to see and thus accurately measure
their BLMAX (in the axial view window). Perform the compen-
satory inclination in the coronal window view as instructed in
Fig. 5.
3. Next, in the sagittal or coronal viewing window, move the axial
plane slightly inferiorly until reaching the level of the tip of the
pulp horns of the tooth (Fig. 3C) (see Note 7).
4. In the axial viewing window, identify the “anatomical”
buccal-most (BMan) point of the tooth’s crown (for details see
Fig. 3A, B) (see Note 8).
5. Next, identify the “anatomical” lingual-most (LMan) point of
the tooth’s crown (for details see Fig. 3A, B) (see Note 8).
6. Lastly, still in the axial view window, use your software mea-
surement tools to measure the M3 crown from its “anatomi-
cal” buccal-most (BMan) point to its “anatomical” lingual-
most (LMan) point (Fig. 3A, B). Provided that any required
compensatory inclination of the axial plane is made, this
method may be used to measure the M1 and M2 crowns as
well (see Note 9).
3.3 The Third Molar The M3 eruption space in the maxilla is most often defined as a
Eruption Space in the linear distance between the distal-most point of the M1 or M2
Maxilla crown and the maxillary tuberosity (T). Although other maxillary
land markers may be used to define the M3 eruption space, the
method presented here uses the maxillary tuberosity as the land-
mark of choice because it is the most immediate physical limit distal
to the M3. This distal physical limit is reinforced by the presence of
the pterygoid process of the sphenoid bone (PT), which fuses with
the maxillary tuberosity (i.e., no M3 is likely to develop beyond this
point). Therefore, our method also takes into account the ptery-
goid process. Here we describe the steps for measuring this maxil-
lary eruption space using CBCT images:
1. Complete Subheading 3.1, steps 1–6 for the maxillary quad-
rant, if you have not already done so. These steps must be
already done before moving through the next steps.
2. Determine whether your study requires the M3 eruption space
to be measured from the M1 or M2. Also, make sure that the
axial plane remains positioned at the contact level (CL) (step 5)
(see Note 10).
3. In the axial window view, move the axial plane superiorly to
the level where the maxillary tuberosity fuses with the
pterygoid process of the sphenoid bone (Fig. 6B) (see

Note 11).
4. Next, mark the contour of the maxillary tuberosity using the
particular tools of your software package (Fig. 6B). This con-
tour line demarks the distal limit of the M3 eruption space in
the maxilla (see Note 12).
5. Next, move the axial plane once again down to the level of
CL. The axial plane needs to be moved back to CL because that
is the level at which the axial plane remains while all the mea-
surements are taken.
6. Subsequently, using your software marking tools (e.g., lines,
arrows), determine the following set of mesiodistal lines to
guide your future linear measurements (Fig. 6C) (see Note 13):
(a) Buccal guiding line (B): This line should ideally start at the
horizontal mid-third of the buccal surface of the M1 or
M2 (step 2), at its “absolute” buccal-most point (BMab),
and extend distally to the contour of the maxillary tuber-
osity (step 4).
(b) Lingual guiding line (L): This line should ideally start at
the horizontal mid-third of the lingual surface of the M1
or M2 (step 2), at its “absolute” lingual-most point
(LMab), and extend distally to the contour of the maxillary
tuberosity (step 4).
(c) The above two lines should be placed parallel to each
other while also “englobing” at its best the M2 in between
them. If necessary, make adjustments at this moment,
always preserving the parallelism between the two lines.
7. Identify the “absolute” distal-most point (DMab) of the M1 or
M2 crown (Fig. 6C) (see Note 14).
8. The length of the M3 eruption space in the maxilla can now be
measured mesiodistally (see Note 15), as follows:
(a) Use the measurement tools of your software to measure
the distance between the “absolute” distal-most point
(DMab) of the M1 or M2 crown and the contour of the
maxillary tuberosity (Fig. 6C).
(b) For reproduction of the method, this linear measurement
should be done parallel to the buccal and lingual guiding
lines, respectively, determined in steps 6(a) and 6(b)
(of this section).
9. As noted earlier, we recommend recording the measurements
in word processing or spreadsheet software, especially if your
CBCT viewing software does not allow you to save your work
and/or export your data for analysis (see Note 16).
Fig. 6 Measuring the M3 eruption space in the maxilla and mandible. (A) The axial plane can be moved along
various superior-inferior levels along the sagittal view of an individual’s CBCT image. Each level, represented
by slices B–D, generates a specific axial image. Blue line: occlusal plane level for upper molars. Red line:
occlusal plane level for lower molars. (B) At this level the maxillary tuberosity (T) fuses with the pterygoid
process (PT) of the sphenoid bone. The more superior you move your axial plane relative to the occlusal plane
of upper molars (blue line), the more evident become the fusion between these two structures (T and PT). For
measurement consistency, we recommend using the level where this fusion is first seen. At this level the
distal contour of the maxilla should be highlighted using your software marking tools (e.g., lines, arrows), as
done in B above. (C) When the axial plane is moved inferiorly to the level of the interproximal contact points of
the M1 and M2 (contact level or CL), both molar crowns are seen at their widest mesiodistal dimensions. Mark
buccal (B) and lingual (L) lines to guide your measurements. Next, identify the real distal-most point (DMab) of
the M1 or M2 crown. We chose the M1 to demonstrate our measurement method. Then, use your software
measurement tools to measure the distance from DMab to the distal limit of the maxillary tuberosity, done
parallel to the buccal and lingual guiding lines. Note that the distal limits of the maxilla may be determined
3.4 The Third Molar The M3 eruption space in the mandible is most often defined as a
Eruption Space in the linear distance between the distal-most point of the M1 or M2
Mandible crown and the ascending ramus of the mandible (R). Although
other mandibular landmarks (e.g., mandibular foramen) may be
used to determine the length of M3 eruption space, our method
uses the anterior edge of the ascending ramus as the landmark of
choice because it is the most immediate physical limit distal to the
M3. Here we describe the steps for measuring M3 eruption space in
the mandible using CBCT images:
1. If Subheading 3.1, steps 1–6, is already done for the mandib-
ular quadrant under investigation, then proceed with the next
instructions. These steps must be already done before moving
through the next steps.
2. Determine whether your study requires the M3 eruption space
to be measured from the M1 or M2. Also, make sure that the
axial plane remains positioned at the contact level (CL) (step 5)
(see Note 10).
3. In the axial viewing window, determine the mesial-most limits
of the buccal (BC) and lingual (LC) bone corticales of the
mandible. Use your software’s markers (e.g., lines, arrows) to
identify these two specific points (Fig. 6D) (see Note 17).
4. Trace a line connecting the previously determined two points
(BC, LC) (Fig. 6D). This line demarcates the distal limit of the
M3 eruption space in the mandible. This is also the anterior-
most limit of the mandibular ascending ramus (R) at this
superior-inferior level (see Note 18).
5. Subsequently, using your software marking tools (e.g., lines,
arrows), determine the following set of mesiodistal lines to
guide your future linear measurements (Fig. 6D) (see Note
13):
(a) Buccal guiding line (B): This line should ideally initiate at
the horizontal mid-third of the buccal surface of the M1
or M2 (step 2) at its “absolute” buccal-most point
(DMab) and extended distally until reaching the anterior
edge of the ramus (step 4).
Fig. 6 (continued) either by marking the area with your software tools (preferably) or by using the coronal line
(green line) as parameter. (D) When the axial plane is positioned at the level of the interproximal contact points
of the M1 and M2 (contact level or CL), the molar crowns are seen at their widest mesiodistal dimensions.
Mark buccal (B) and lingual (L) lines to guide your measurements. Next, identify the real distal-most point
(DMab) of the M1 or M2 crown. We chose the M1 to demonstrate our measurement method in this image.
Then, identify the anterior-most points of the buccal (BC) and lingual (LC) mandibular corticales. A line should
be drawn connecting these two points. This line shows the location of the anterior-most limit of the ascending
mandibular ramus (R). Then, use your software measurement tools to measure the distance from DMab to the
line R. This measurement should be done parallel to the buccal and lingual guiding lines
(b) Lingual guiding line (L): This line should ideally start at
the horizontal mid-third of the lingual surface of the M1
or M2 (step 2) at its lingual-most point and extend
distally until reaching the anterior edge of the ramus
(step 4).
(c) The above two lines should be parallel to each other while
also “englobing” at its best the M2 in between the two
parallel lines. If necessary, make adjustments at this
moment, always preserving the parallelism between the
two lines.
6. Identify the “absolute” distal-most point (DMab) of the M1 or
M2 crown (see Note 14).
7. The length of the M3 eruption space in the mandible can now
be linearly measured mesiodistally (see Note 15), as follows:
(a) Use the measurement tools of your software to measure
the distance between the “absolute” distal-most point
(DMab) of the M1 or M2 crown (step 2) and the line
determined in step 4 (representing the ascending man-
dibular ramus) (Fig. 6D).
(b) For reproducibility, this measurement should be done
parallel to the buccal and lingual guiding lines, respec-
tively, determined in steps 5(a) and 5(b) (of this section).
8. Again, be sure to have a robust way to record, save, and export
your measurements for subsequent analyses (see Note 16).
4 Notes
1. Make sure that your imaging files are saved in a format com-
patible with the imaging software chosen. Also, make also sure
to assess the individual CBCT’s image for inclusion or exclu-
sion factors. You may want to exclude from the study, for
instance, individuals presenting severe dental crowding
and/or malocclusions, as well as those with osseous craniofacial
defects or with artificial gaps in the dental arch as a result of
previously extracted permanent teeth. Adhering of these exclu-
sion criteria is especially important if your study investigates the
third molar eruption space in the jaws.
2. The thinner the slice, the more precise the analysis of targeted
anatomical structures. Thick slices may generate images with
one or more anatomical structures superimposed, making
assessment of the M3 and its eruption area more challenging.
3. The coronal and sagittal planes should intersect each other
right on the center of the M1’s crown so that the M1 (which
is a reference tooth in our method) is also shown in the coronal
and sagittal window views. Although the structure under inves-

tigation is the M3 (and its eruption area), tilting of the M1’s
long axis occurs relatively less frequently, and, for this reason, at
some moments we need to use the M1 as a reference structure
(e.g., to configure the spatial position of the individual’s head
and jaws).
4. Finding the occlusal level is necessary to guide us through the
next steps of our method. Always use the M1 as your most
“stable” or “reliable” reference. The M2 and M3 tend to be
inclined along their long axis more frequently than the M1.
Ideally, the M1 and M2 should be at the same occlusal level in
order to use the M3 measurement method presented here.
However, in some individuals, the M2 may be at a slightly
inferior level relative to the M1 (especially in teenagers). In
such cases, determine the occlusal plane for that oral quadrant
over the M1 cups tips, extending to premolars or incisors. If the
M1 is not yet fully erupted, consider excluding the individual
from the study.
5. The maximum mesiodistal diameter of a molar is more fre-
quently found at the superior-inferior crown level where this
molar touches its adjacent teeth. That is, in most cases the
molar’s contact areas (to adjacent teeth) are within the
mid-third of its crown (superior-inferiorly). Therefore, for a
more accurate picture of the M1’s maximum mesiodistal crown
diameter, use the sagittal window view to determine the level of
the M1 contact areas.
6. Always observe the angulations of the M3’s long axis relative to
its neighbors’ long axes, the M1 and M2. Note that the M1 is
found in a vertical position relatively more often than the M2.
For this reason, the position of the M1’s long axis should be
considered as a more “stable” or “reliable” reference compared
to the M2.
7. The maximum buccolingual diameter of a molar crown can be
seen and measured, in most cases, at the level of the tooth’s
pulp horn tips—not at the level of the contact point level,
which are located slightly above in the tooth’s crown. Varia-
tions in dental anatomy may, of course, occur and need to be
accounted for, especially when the structure under investiga-
tion is the M3, which has a highly variable morphology. Thus,
make sure that no significant deviations from normality exist in
the crown’s anatomy and that the crown’s maximum buccolin-
gual diameter is indeed at this level (e.g., level of the tooth’s
pulp horn tips).
8. For illustrative details on how to identify the anatomical and
absolute extreme points of a tooth’s crown, please refer back to
Fig. 3. Note that the location of anatomical and absolute points
may not coincide. The “anatomical” buccal-most point (BMan)

of a tooth’s crown, for instance, will be always located on the
crown’s buccal surface, even if the tooth is rotated (e.g., gir-
overted) to a new position (e.g., mesially or distally). There-
fore, anatomically speaking, the crown’s buccal surface will
remain a “buccal surface” even after rotation (see Fig. 3 for
further details).
9. Note that, for consistent measurements, the maximum mesio-
distal diameter of a tooth’s crown is always represented by the
distance between this crown’s anatomical mesial-most (MMan)
and anatomical distal-most (DMan) points, and not by the
space actually taken by that tooth in that specific dental arch.
A tooth crown with a mesiodistal diameter of 12 mm, for
instance, will always measure 12 mm even if it is giroverted or
tilted (see Fig. 3 for further details). Whether or not the tooth’s
long axis is tilted, always consider measuring its maximum
mesiodistal and buccolingual dimensions on the axial window
view. Relative to sagittal and coronal images, axial images of
anatomical structures are subject to less superimposition by
adjacent structures (which occur especially if thick image slices
are used, e.g., >10 mm) and for this reason are preferred for
taking measurements.
10. Your study may use the M1 or M2 as a landmark to identify the
mesial limit of the M3’s eruption space. Consider using the M2
if you are studying adults (e.g., age 17+). In such individuals
the M2 is likely to have already erupted to the same occlusal
level as the M1. However, to study individuals age 12 or youn-
ger, consider using the M1 as a landmark. These individuals’
M2s may present variable levels of eruption, which makes the
M2 an “unreliable” landmark for measurement purposes.
11. Because the pterygoid process of the sphenoid bone fuses with
the maxillary tuberosity (e.g., approximately at the level of the
M1 and M2 root apices), no M3 is likely to develop beyond this
point, which makes this point an important landmark to set the
distal limits of the space available for M3 eruption.
12. The contour of the maxillary tuberosity may be marked easily
using your software marking tools (e.g., lines, arrows). These
markings are intended to guide your measurements and, there-
fore, should ideally remain visible, when you move the axial
plane inferiorly to the level of the contact point’s level of the
M1 and M2. Inspect your software marking tools to determine
which one(s) allow the above task to be performed. If your
tools do not allow such task to be performed, use the coronal
plane (in the axial view window) to determine the location
where the pterygoid process fuses with the maxilla.
13. The presence of these guiding lines is not strictly necessary.

However, in our experience, they improve the reliability of our
measurements (e.g., inter-examiner measurement results). For
this reason, we highly recommend using these lines.
14. Note that now, different from step 9, we seek to measure the
actual space available for M3 eruption (not the size of the M3
crown). Therefore, at this moment we need to identify the
“absolute” distal-most point (DMab) reached by the crown of
the M1 or M2 in the oral quadrant under investigation (not the
“anatomical” distal-most point (DMan)). Thus, careful obser-
vation is necessary, especially if the M1 or M2 is giroverted or
tilted.
15. Provided that all the previous steps have been carefully
observed, measuring the M3’s eruption space itself should
not be a challenging task. In order to assure the consistency
of your measurement results, however, we recommend doing a
set of initial measurements at regular time intervals, done by at
least two investigators, if possible, and calculating the standard
deviation and error between two or more sets of metrics.
16. Store your data in a safe location (e.g., external hard drive).
Perform regular data backups. Whenever possible, create
password-protected storage files (particularly important for
working with patient data, even if de-identified).
17. The buccal and lingual mandibular osseous corticales are easily
identified on CBCT images. Look for a thick radiopaque line in
the peripheral areas of the mandible. A thin image slice (e.g.,
1 mm or less) is especially important during this step: a thin
slice will provide a sharper or more accurate view of the
anterior-most edges of these corticales. For details, refer back
to Subheading 3.1, step 2.
18. This line lays buccolingually in most individuals, but not nec-
essarily perpendicular to the curved mesiodistal line that con-
nects all teeth in that dental arch. For illustrative details refer to
Fig. 6.
Acknowledgments
The authors thank the Colleges of Dentistry and Medicine, the

Department of Anatomy and Cell Biology, and the University of
Saskatchewan, for allowing the use of specialized facilities as well as
for the generous technical and financial resources provided in sup-
port to this project. Without their kind support, the development
of the present study methodology would not be possible.
References
1. Liversidge HM (2008) Timing of human man- 6. Dudhia R, Monsour PA, Savage NW, Wilson RJ
dibular third molar formation. Ann Hum Biol (2011) Accuracy of angular measurements and
35(3):294–321 assessment of distortion in the mandibular third
2. Mohammed B, Mansur A (2013) Relationship molar region on panoramic radiographs. Oral
of the inferior alveolar canal to impacted third Surg Oral Med Oral Pathol Oral Radiol Endod
molars as evaluated by cone beam computed 111(4):508–516
tomography. Northwest Dent 92:35–37 7. Marchiori DF, Packota GV, Boughner JC
3. Suomalainen A, Venta I, Mattila M, Turtola L, (2016) Third-molar mineralization as a function
Vehmas T, Peltola JS (2010) Reliability of CBCT of available retromolar space. Acta Odontol
and other radiographic methods in preoperative Scand 74(7):509–517
evaluation of lower third molars. Oral Surg Oral 8. Ghougassian SS, Ghafari JG. Association
Med Oral Pathol Oral Radiol Endod 109 between mandibular third molar formation and
(2):276–284 retromolar space. Angle Orthod [Internet].
4. Nguyen E, Boychuk D, Orellana M (2011) 2014 Apr 28:1–5 p. http://www.ncbi.nlm.nih.
Accuracy of cone-beam computed tomography gov/pubmed/24773221
in predicting the diameter of unerupted teeth. 9. Lagravère MO, Carey J, Toogood RW, Major
Am J Orthod Dentofac Orthop 140(2):e59–e66 PW (2008) Three-dimensional accuracy of mea-
5. Sonick M, Abrahams J, Faiella RA (1994) A surements made with software on cone-beam
comparison of the accuracy of periapical, pano- computed tomography images. Am J Orthod
ramic, and computerized tomographic radio- Dentofac Orthop 134(1):112–116
graphs in locating the mandibular canal. Int J
Oral Maxillofac Implants 9(4):455–460
Chapter 32
Protocols to Study Dental Caries In Vitro: Microbial

Caries Models
Bennett T. Amaechi, Livia M. A. Tenuta, Antonio P. Ricomini Filho,
and Jaime A. Cury
Abstract
Caries lesions result from the interaction between dental biofilm and sugars. Since the biofilm is an
important component in the etiology of the disease, biofilm models have been developed to study the
cariogenicity of dietary sugars, as well as the anticaries effect of substances. Two of such models, termed as
“static” or “continuous flow,” are described in details here together with their advantages, limitations, and
applications.
Key words Dental caries, Dental plaque, Microbial model, Static model, Artificial mouth, Continu-
ous flow model
1 Introduction
Dental caries lesions develop as a result of the metabolic action of a

microbial biofilm adhered on teeth, when it is frequently exposed to
fermentable sugars. Therefore, caries can be modeled by using
microbial models, in which a cariogenic biofilm is formed onto
hard tissue specimens (enamel, dentin, hydroxyapatite), under
exposure to fermentable sugars.
The advantages of microbial models, over other in vitro caries
models such as pH-cycling models, are that the antibacterial effect
of anticaries agents can be studied. They also allow the study of the
cariogenic properties of the biofilms, such as its composition
(microbial and biochemical) and structure (spatial arrangement of
the bacteria and extracellular matrix). In addition, the metabolism
of the biofilm, including the expression of genes, can be explored.
Most, if not all, microbial models described so far are used to assess
demineralization (or inhibition of demineralization) of dental
mineralized tissues. Given that the biofilm per se does not have
357
358 Bennett T. Amaechi et al.
properties to induce dental remineralization (it is the surrounding

fluid composition, such as the mineral ions concentration in saliva
and biofilm fluid, which induces remineralization), apparently there
is no reason to develop a remineralizing microbial model (unless for
the study of calculus formation in the biofilm).
There are many types of microbial caries models described in
the literature. They can be classified according to different condi-
tions, such as the number of bacterial species used (single versus
multispecies), the length of the model (short-term versus long-
term), the type of exposure to sugar (continuous versus intermit-
tent), or the type of flow of culture medium (static or continuous).
In the current chapter, we will describe two types of microbial caries
models, classified according to their main characteristics.
2 Static Model
The model described here was proposed by Ccahuana-Vasquez and

Cury [1], and it presents the following advantages over other static
biofilm models: (a) it was validated for the dose effect of chlorhexi-
dine, the gold standard antimicrobial substance to control oral
biofilms; (b) it allows the use of enamel (or dentin) slabs in order
to assess mineral loss; (c) sugar exposure is intermittent (instead of
continuous), in order to mimic the feast and famine episodes to
which the biofilm is exposed to in the oral cavity, allowing for a
more reliable estimate of the biofilm properties (microbial popula-
tion, extracellular matrix composition, and gene expression); and
(d) it uses a suspended substrate design (instead of bottom lying) to
allow for true adhesion of the bacteria during the first phases of
biofilm formation, determination of the pH (indicator of biofilm
acidogenicity), concentration of released calcium (indicator of min-
eral loss), and bacterial metabolic products in the culture medium
used in the model. The model was further validated for the dose
effect of fluoride on the reduction of enamel and dentin demineral-
ization [2] and has been successfully used to evaluate the antimi-
crobial effect of iron [3] and other natural products [4], as well as to
evaluate the cariogenic potential of other dietary carbohydrates
[5–8]. A modified version of this model was used to evaluate the
effect of antimicrobial toothpaste [9]. In addition, the model can
be used to evaluate properties of the extracellular matrix [10].
Although this is a static model, i.e., the culture media are not
flowing continuously over the growing biofilm, the media are
continuously changed during the experiment, and the cariogenic
challenge is not continuous, it is intermittent. Therefore, it can be
regarded as a cycling model. Furthermore, the medium pH and
calcium concentration are correlated with the demineralization that
occurred in enamel and dentin [2, 8, 10].
Microbial Models for Studying Dental Caries 359
2.1 Materials 1. Cut 4 mm 7 mm enamel (or dentin) slabs from bovine

incisor crowns (or roots). Abrade the inner (dentin in case of
2.1.1 Enamel (or Dentin)
enamel) surface of the slabs using a polishing machine until a
Slabs
thickness of about 1.2–1.5 mm is reached.
2. Polish the surface of enamel (or dentin) with 400-grit,
600-grit, and 1200-grit papers, followed by a 1 μm diamond
suspension-embedded polishing cloth to obtain flat, scratches-
free surfaces.
3. Baseline slab analysis: Decide your measurement method(s).
Two common standard measurements are (1) microhardness
(surface [SMH] or cross-sectional [CSMH]) and (2) transverse
microradiography (TMR). Due to the short-term nature of the
model, and depth of the caries lesions, surface microhardness
may be the most suitable method to determine differences
among the groups in test. However, to evaluate enamel-dentin
demineralization by CSMH or TMR, it is mandatory to
increase the duration of biofilm treatment with sucrose or to
use the feeding model with 1% sucrose constantly present in the
media (see protocol below).
(a) Measure the surface microhardness of the blocks in a
microhardness tester, with a Knoop diamond indenter, at
a 50-g load for enamel and 5-g load for dentin, for 5 s.
Three to five indentations should be made in the center of
the slab, and averaged.
(b) Exclude slabs with very low or very high surface hardness
values (e.g., select slabs within a range of 10–20% above or
below the average hardness value of all slabs) and those
with a great variability among the indentation values (e.g.,
exclude slabs with a coefficient of variation of the 3–5
indentations greater than 10%).
2.2 Methods 1. Prepare orthodontic wire holders to maintain the slabs in a

vertical position inside the 24-well plate (Fig. 1).
2.2.1 Preparation of the
24-Well Plates for Biofilm 2. Mount a 24-well plate with the enamel (or dentin) slabs on the
Formation holders in each well, and sterilize it using ethylene oxide or
gamma radiation.
2.2.2 Bacteria and 1. You can choose to develop and use single or multispecies
Culture Media biofilm. For single-species biofilm, Streptococcus mutans
(UA159, ATCC 700610) is the bacteria of choice for demin-
eralization experiment. However, models using S. mutans
mixed to other species (multispecies biofilm) can be developed
if there is interest for studying starchy dietary products [7] or
different bacterial interactions [11].
Fig. 1 Enamel slab mounted in the metallic holder (a) to maintain the slab in a suspended vertical position
inside the well of a 24-well plate (b, c)
2. The choice of the amount of inoculum will depend on the

growth rate and interaction among the species used in the
model (pilot studies may be required).
3. Use a phosphate-buffered medium (e.g., tryptone-yeast extract
broth; pH 7.0) with low glucose concentration (0.1 mM; sali-
vary basal concentration) in order to avoid continuous expo-
sure of the slabs to demineralizing conditions.
4. Saliva may be added to the culture media in case starchy pro-
ducts will be tested in the model (e.g., saliva is needed for the
degradation of starch by amylase when testing the cariogenic
potential of starchy products) [8].
2.2.3 Biofilm Formation 1. Treat slabs with saliva in order to form acquired pellicle.
and Cariogenic Challenges (a) Saliva can be previously collected on ice from donors
chewing paraffin film, after at least 2 h of fasting, mixed
with adsorption buffer (1:1) containing a protease inhibi-
tor [12] and then centrifuged at 3800 g for 10 min at
4 C. Collect supernatants and filter using a 0.2 μm filter.
Filtered saliva can be stored on ice until use.
(b) Load the 24-well plate containing the slabs with filtered
saliva and maintain at 37 C for 30 min, under agitation
(60 rpm, orbital shaker).
2. Transfer the slabs from saliva to the culture media containing
the bacterium (a) inoculum, and maintain for 6–8 h at 37 C,
and appropriate atmosphere, for initial adhesion. The medium
used for bacterial adhesion usually contains 1% sucrose to
enhance the adhesion of cariogenic bacteria (i.e., S. mutans).
You may also need to increase the buffer capacity (e.g., 10
higher than the usual concentration) in the culture media to
avoid demineralization during the adhesion phase [10].
Fig. 2 Enamel slab immersed in culture medium (a, b) during the biofilm formation. Only adhered bacterial
cells on the enamel surface will grow to form biofilm (c)
3. After the initial adhesion, change the culture medium to a fresh

one, and incubate overnight.
4. On the second day, first change the culture medium, and then
start the exposure to cariogenic challenges. Expose the slabs to
a sugar solution (of the chosen cariogenic sugar) for 3 min,
eight times per day (from morning till late afternoon), simulat-
ing a high cariogenic challenge.
5. After every cariogenic challenge, rinse each slab in 0.9% NaCl
solution, and return to the same culture media (Fig. 2).
6. If anticaries treatments are being used, alternate them with the
eight cariogenic challenges of the day. You may want to test
fluoride rinses or chlorhexidine rinses, as controls. Use the
frequency most likely recommended for in vivo use (e.g.,
1 min, twice/day exposure to the treatment) [1–3, 9]. Rinse
the slabs in 0.9% NaCl solution before returning to the culture
media.
7. After the last cariogenic challenge of the day, change again the
culture media. During the night, no cariogenic challenges are
performed.
8. Repeat the steps previously described (items 4–7) as necessary.
9. On the 4th (for dentin) or 5th (for enamel) days, the experi-
ment is terminated, and the biofilm and slabs are collected for
analysis.
2.2.4 Culture Media 1. Measure the pH of the culture media at every medium change.
Analysis (Fig. 3) The pH of the medium in which the slabs were kept during the
day (time of the cariogenic challenges) should have decreased;
the pH of the culture media in which the slabs were kept
overnight should be close to the baseline pH.
2. The concentration of calcium in the culture media can be
determined and used as a surrogate measure of tooth
demineralization [2].
Fig. 3 Flowchart of the static model protocol including the steps to form the cariogenic biofilm and the
suggested analyses for biofilm, enamel slab, and culture medium
3. Bacterial metabolic products, such as acids produced due to the

metabolization of sugars, can also be identified and quantified
in the culture medium [10].
4. Fluoride can also be determined in the culture media if fluoride
has been used in the experiment [2].
2.2.5 Biofilm Analysis 1. Remove each slab from the holder using a sterile plier, and
immerse it in 1 mL of sterile 0.9% NaCl solution. Sonicate at
7 W for 30 s to detach the biofilms from the slabs. The bacterial
suspension can be used for different analyses:
(a) Determine viable bacteria by plating the suspension in
appropriate culture media to quantify bacterial population
in biofilms.
(b) Determine protein amount in the suspension using a col-
orimetric method (e.g., Lowry and bicinchoninic acid—
BCA) to estimate the bacterial proportion in biofilms.
(c) Determine the biofilm dry weight by drying an aliquot of
the suspension to quantify the biofilm biomass (bacterial
cells plus matrix).
(d) Determine the concentration of soluble and insoluble
extracellular polysaccharides using a colorimetric method
for total carbohydrate analysis (phenol-sulfuric method)
to quantify the amount of extracellular polysaccharide
present in the matrix and to estimate the proportion of
each type (soluble and insoluble).
(e) In case the biofilm (bacterial cells and/or matrix) is to be

visualized using microscopy techniques (e.g., confocal
laser scanning microscopy, CLSM, and transmission elec-
tron microscopy, TEM), collect the slabs with the intact
biofilm and proceed to specific protocol required previ-
ously to microscopy visualization.
2.2.6 Posttreatment Slab 1. Collect the slab from the saline solution and clean with a soft
Analysis tissue.
2. Analyze the slabs for demineralization, and measure mineral
loss and lesion depth using one or a combination of any of the
following:
(a) Determine the posttreatment surface and/or cross-
sectional microhardness, and calculate the percentage
loss in surface microhardness [13] or the area of caries
lesions [14, 15].
(b) Perform posttreatment TMR analysis to determine the
mineral loss and lesion depth [16, 17].
(c) Verify demineralization, and measure the depth of demin-
eralization using polarized light microscopy [18, 19].
2.2.7 Data Analysis 1. Replicate the experiment at least three times on different occa-
sions to check for reproducibility.
2. Analyze the data for the independent experiments, according
to the recommended statistical tests and the groups being
compared.
3 Continuous Flow Model
The continuous flow microbial model described here was pro-

posed by Amaechi and his cariology research group and was vali-
dated for the dose effect of chlorhexidine, the gold standard
antimicrobial substance to control oral biofilms [20]. It has all
the characteristics of the static model described above except that
the culture media flow continuously over the substrate and the
growing biofilm to simulate the oral fluid, saliva. However, cario-
genic challenge and the treatment with test products such as
toothpaste or mouthrinse are intermittent. The model was further
validated for simulation of caries process [21] and the antimicro-
bial effect of a mouthrinse [22]. It has also been successfully used
to evaluate the effect of an anticaries agent on the reduction of
enamel demineralization [17].
3.1 Materials Sound human or bovine teeth are collected and cleansed of soft
tissue debris, brushed with pumice slurry using an electric tooth-
3.1.1 Sample
brush, and then examined by transillumination. Teeth without
Preparation
cracks, hypoplasia, white spot lesions, and other malformations
are selected. Substrate can be whole tooth or enamel block or
dentin blocks. The sample analysis method should determine the
specimen preparation.
3.1.2 Sample Size Each of the selected teeth or tooth block is randomly assigned to
the experimental treatment groups, with a minimum of ten teeth
per group. The number of specimens must be sufficient to support
statistical separation of positive and negative controls and enable a
sufficient power to detect desired differences.
3.2 Methods If the test products are toothpaste, freshly prepared slurries of
both test and control toothpaste are made with sterilized deio-
3.2.1 Test Product
nized distilled water. A dilution of one part toothpaste to three
Preparation
parts water is recommended, as this represents the anticipated
level of dilution that occurs during routine use of toothpaste
products. Thoroughly mix for 4 min, using a laboratory stand
mixer until homogenous; 4.0 mL of the 1:3 with diluent is used
per tooth.
3.2.2 Bacteria and You can choose to develop and use single or multispecies biofilm.
Culture Media For single-species biofilm, Streptococcus mutans (UA159, ATCC
700610) is the bacteria of choice for demineralization experiment.
For multispecies in demineralization experiment, Streptococcus
mutans, Lactobacillus casei, and Actinomyces viscosus are preferable
bacteria. The choice of culture media depends on the type of
culture (single or mixed) and type of bacteria.
3.2.3 System Description This system is composed of multiple-station continuous flow cul-
ture chambers (Fig. 4), which acts as an artificial mouth and simu-
lates the biological and physiological activities observed within the
oral environment. Each station consists of a chamber bearing (1) a
cylindrical clear-acrylic rod with vertical grooves for mounting
either whole tooth or tooth blocks; (2) a head assembly with
three lines for supply of simulated oral fluid (culture media), nutri-
ents, experimental reagents, and inoculation of the chamber with
either single- or multispecies bacterial consortium; and (3) access
for plaque sampling and electrode insertion for pH monitoring. All
components of the system are sterilized using ethylene oxide gas
and aseptically set up. The simulated oral fluid (SOF) used in this
system depends on the type(s) of microorganisms being used; the
commonly used media is Bacto™ Todd Hewitt Broth (since it
supports multiple organisms) with pH adjusted to 7.0. This is
continuously circulated to simulate saliva. Continuous circulation
through the chambers at individually controlled flow rates
Fig. 4 Schematic representation of the artificial mouth system and its

components. (a) Programmable circulating pump for broth; (b) programmable
pump for broth and sucrose; (c, d, e) oral chambers; F, Todd Hewitt broth; G,
sucrose; H, return-flow line; I, tooth block; J, broth and sucrose pumping
tubes; K, broth circulating tube
(0.3 mL/min) via digital programmable pumps is maintained from

a reservoir. A complete circulatory system is established by a return-
flow line from the chamber back into the reservoir. The reservoir
content is changed daily. The flow rate of the SOF is varied in
accordance with the oral condition being simulated (e.g., stimu-
lated or unstimulated salivary flow). Sugar exposure is intermittent,
in order to mimic the feast and famine episodes to which the
biofilm is exposed to in the oral cavity. Thus 10% sucrose is supplied
(flow rate 0.7 mL/min) three times daily for 6 min on each occa-
sion to simulate meals. All fluids flow uniformly as a thin film over
the surface of the rod. The entire assembly is housed inside a reach-
in CO2 incubator maintained at 5% CO2 and at a constant physio-
logical temperature of 37 C. A micro-esophageal glass pH elec-
trode and micro-reference electrode connected through a pH
meter are installed in each chamber at the plaque growth surface
to monitor the intra-plaque pH.
3.2.4 Treatment 1. The experimental groups are randomly assigned to the culture
Regimen chambers in the “artificial mouth system.” Using heavy-duty
putty, the specimens are embedded in the vertical grooves on
the surface of the cylindrical rod in the culture chamber. The
specimens are embedded such that their surfaces flushed with
the surface of the cylinder to permit streamlined flow of fluids,
and the exposed specimens are available for plaque growth.
2. The system is operated as described above by continuous circu-
lation of the chosen culture media separately through each
chamber to simulate saliva flow, and 10% sucrose is supplied
three times daily for 6 min on each occasion to simulate meals
and pH cycling. The pH of plaque in each chamber is moni-
tored at nonfeeding time to check maintenance of neutrality by
CO2.
3. To initiate biofilm growth and caries development on the
experimental tooth specimen, culture inoculated with either
single or multispecies (broth to inoculum ratio 10:1) is circu-
lated through the chamber for 12 h on day 1 (adhesion phase).
Then bacteria-free broth is circulated for the rest 12 h of day 1.
4. From day 2, the plaque-covered tooth blocks are treated as
shown in Table 1. Briefly, the control group receives no treat-
ment, while the test groups are treated with their respective
products (toothpaste slurries or mouthrinses) once daily for
1 min (if mouthrinses) or twice daily (morning and evening)
for 2 min (if toothpaste) or according to the study directives,
on each occasion as follows. The cylindrical rod bearing the
tooth specimen is immersed into the product for the specified
Table 1
Treatment schedule for artificial mouth system for this study
Day Time Treatment

Day 1 8:00 Circulation of bacteria-free culture media starts
10:00–11:00 Bacteria-inoculated media is circulated for 12 h (adhesion phase)
11:00 Circulation of bacteria-free media restarts
20:00 Sucrose circulation for 6 min
20:06 till next morning Circulation of bacteria-free media restarts
Day 2–Day 7 7:00 Toothpaste (2 min) or mouthrinse (1 min) treatment
19:00 Toothpaste (2 min) or mouthrinse (1 min) treatment
period of time and then gently rinsed with sterile phosphate

buffer saline (PBS). For toothpaste, fresh slurry of each tooth-
paste sample is prepared just prior to each treatment episode,
and the pH of the toothpaste slurry is measured before
treatment.
5. All treatments are carried out inside the incubator at 37 C and
under aseptic condition.
6. The experiment lasts from 7 (dentin) to 14 (enamel) days. On
termination of the experiment, the biofilm and the tooth speci-
mens are harvested for analysis.
3.2.5 Data Management All the analyses for the culture media, biofilm, and tooth slabs
described above for the “static model” protocol can be performed
with the “continuous flow” model.
3.2.6 Data Analysis Analyze the data for the independent experiments, according to the
recommended statistical tests and the groups being compared.
References
1. Ccahuana-Vásquez RA, Cury JA (2010) 7. Cavalcanti YW, Bertolini MM, da Silva WJ,
S. mutans biofilm model to evaluate antimicro- Del-Bel-Cury AA, Tenuta LM, Cury JA
bial substances and enamel demineralization. (2014) A three-species biofilm model for the
Braz Oral Res 24(2):135–141 evaluation of enamel and dentin demineraliza-
2. Fernández CE, Tenuta LM, Cury JA (2016) tion. Biofouling 30(5):579–588
Validation of a cariogenic biofilm model to 8. Botelho JN, Villegas-Salinas M, Troncoso-
evaluate the effect of fluoride on enamel and Gajardo P, Giacaman RA, Cury JA (2016)
root dentine demineralization. PLoS One 11 Enamel and dentine demineralization by a
(1):e0146478 combination of starch and sucrose in a bio-
3. Ribeiro CC, Ccahuana-Vásquez RA, Carmo film—caries model. Braz Oral Res 30(1).
CD, Alves CM, Leitão TJ, Vidotti LR, Cury https://doi.org/10.1590/1807-3107BOR-
JA (2012) The effect of iron on Streptococcus 2016.vol30.0052
mutans biofilm and on enamel demineraliza- 9. Fernández CE, Fontana M, Samarian D, Cury
tion. Braz Oral Res 26(4):300–305 JA, Rickard AH, González-Cabezas C (2016)
4. Peralta SL, Carvalho PHA, Ccahuana-Vásquez Effect of fluoride-containing toothpastes on
RA, Pereira CMP, Cury JA, Piva E, Lund RG enamel demineralization and Streptococcus
(2017) Cytotoxicity, genotoxicity and antibio- mutans biofilm architecture. Caries Res 50
film activity on Streptococcus mutans of an (2):151–158
experimental self-etching adhesive system con- 10. Costa Oliveira BE, Cury JA, Ricomini Filho AP
taining natural Butia capitata oil. Int J Adhes (2017) Biofilm extracellular polysaccharides
Adhes 78:95–101 degradation during starvation and enamel
5. Giacaman RA, Muñoz MJ, Ccahuana-Vasquez demineralization. PLoS One 12(7):e0181168
RA, Muñoz-Sandoval C, Cury JA (2012) 11. Fernández CE, Giacaman RA, Tenuta LM,
Effect of fluoridated milk on enamel and root Cury JA (2015) Effect of the probiotic Lacto-
dentin demineralization evaluated by a biofilm bacillus rhamnosus LB21 on the cariogenicity
caries model. Caries Res 46(5):460–466 of Streptococcus mutans UA159 in a dual-
6. Muñoz-Sandoval C, Muñoz-Cifuentes MJ, species biofilm model. Caries Res 49
Giacaman RA, Ccahuana-Vasquez RA, Cury (6):583–590
JA (2012) Effect of bovine milk on Streptococ- 12. Koo H, Vacca Smith AM, Bowen WH, Rosalen
cus mutans biofilm cariogenic properties and PL, Cury JA, Park YK (2000) Effects of Apis
enamel and dentin demineralization. Pediatr mellifera propolis on the activities of strepto-
Dent 34(7):e197–e201 coccal glucosyltransferases in solution and
adsorbed onto saliva-coated hydroxyapatite. surface prereacted glass-ionomer filler. J Inves-

Caries Res 34(5):418–426 tig Clin Dent 8(4). https://doi.org/10.1111/
13. Cury JA, Rebelo MA, Del Bel Cury AA, Der- jicd.12249
byshire MT, Tabchoury CP (2000) Biochemi- 18. Silvertown JD, Wong BP, Abrams SH, Sivagur-
cal composition and cariogenicity of dental unathan KS, Mathews SM, Amaechi BT (2016)
plaque formed in the presence of sucrose or Comparison of The Canary System and DIAG-
glucose and fructose. Caries Res 34 NOdent for the in vitro detection of caries
(6):491–497 under opaque dental sealants. J Investig Clin
14. Kielbassa AM, Wrbas KT, Schulte-Meriting J, Dent 8(4). https://doi.org/10.1111/jicd.
Hellwig E (1999) Correlation of transversal 12239
microradiography and microhardness on in 19. Silvertown JD, Wong BP, Sivagurunathan KS,
situ induced demineralization in irradiated Abrams SH, Kirkham J, Amaechi BT (2017)
and nonirradiated human dental enamel. Arch Remineralization of natural early caries lesions
Oral Biol 44:243–251 in vitro by P11-4 monitored with photothermal
15. Mensinkai PK, Ccahuana-Vasquez RA, radiometry and luminescence. J Investig Clin
Chedjieu I, Amaechi BT, Mackey AC, Walker Dent 8(4). https://doi.org/10.1111/jicd.
TJ, Blanken DD, Karlinsey RL (2012) In situ 12257
remineralization of white-spot enamel lesions 20. Lee VA, Karthikeyan R, Rawls HR, Amaechi
by 500 and 1,100 ppm F dentifrices. Clin Oral BT (2010) Anti-cariogenic effect of a cetylpyr-
Investig 16(4):1007–1014. https://doi.org/ idinium chloride-containing nanoemulsion. J
10.1007/s00784-011-0591-2 Dent 38(9):742–749
16. De Josselin de Jong E, Ten Bosch JJ, Noord- 21. Amaechi BT, Ramalingam K (2014) Evaluation
man J (1987) Optimised microcomputer of fluorescence imaging with reflectance
guided quantitative microradiography on den- enhancement for caries detection. Am J Dent
tal mineralised tissue slices. Phys Med Biol 27:112–116
32:887–899 22. Karthikeyan R, Amaechi BT, Rawls HR, Lee
17. Amaechi BT, Key MC, Balu S, Okoye LO, VA (2011) Antimicrobial activity of nanoemul-
Gakunga PT (2016) Evaluation of the caries- sion on cariogenic Streptococcus mutans. Arch
preventive effect of toothpaste containing Oral Biol 56(5):437–445
Chapter 33
In Vitro Caries Models for the Assessment of Novel

Restorative Materials
Basma Sulaiman Ghandourah, Anna Lefkelidou, Raed Said,
Xanthippi Chatzistavrou, Susan Flannagan, Carlos Gonzáles-Cabezas,
Christopher J. Fenno, Li Zheng, Silvana Papagerakis,
and Petros Papagerakis
Abstract
Due to the high failure rates of traditional dental restorations, there is an ongoing effort to develop
modified and new restorative biomaterials in dentistry. Being the most commonly used restorative material,
most of these efforts primarily aim to improve dental composite. Generally, the main objective of such
modifications is to enhance the restorative physical and antimicrobial properties in order to limit micro-
leakage and inhibit bacterial biofilm cultivation. Herein, we describe the process of designing a simple
in vitro model to assess the physical and antimicrobial properties of novel restorative materials in addition to
evaluating their effect on the fragile balance between enamel de- and remineralization.
Key words Dental caries, Composite resin, Biofilm, Bovine teeth, Enamel
1 Introduction
Approximately, two-thirds of all restorative dentistry involves the

replacement of failed restorations [1, 2]. This cycle of re-restoration
leads to larger restorations, weaker teeth, and increased potential
for more complex treatment needs [3]. Generally, composite resin
is the most widely used restorative material due to its superior
esthetics and its quick setting and handling process [4]. Composite
resin was introduced in the 1960s by Dr. Rafael Bowen and was
widely accepted due to its good physical properties and improved
esthetics [5]. They are now considered the most commonly used
direct restorative material for permanent teeth [6, 7]. Despite their
long-term clinical success, composite restorations have relatively
Basma Sulaiman Ghandourah and Anna Lefkelidou contributed equally to this work
369
370 Basma Sulaiman Ghandourah et al.
high failure rates due to secondary caries and micro-leakage, usually

as a result of polymerization shrinkage, which can lead to weaken-
ing of enamel around the restoration, marginal breakdown, and
subsequent loss of the restoration [8–11]. Indeed, the annual
failure rates range between 1 and 3% [12], but the total failure
rate can reach as high as 24.1% [13] or 31% according to the most
recent systematic review [14].
To address these problems, several approaches have been used
to augment the physical and antibacterial properties of dental
restorations in general and composite in particular. These include
adding antimicrobial materials such as fluoride and chlorhexidine in
addition to the incorporation of quaternary ammonium and metal
particle additives to the resin matrix. The incorporation of these
additives has been proven to enhance the bond strength and the
physical properties of the modified composites [15, 16]. Moreover,
silver and zinc oxide were some of the successfully incorporated
additives that resulted in a decrease in Streptococcus mutans and
Lactobacillus count and therefore reducing recurrent caries [17].
This could be attributed to the silver’s ability to bind to bacterial
DNA and especially to thiol groups (R-SH), which inhibits bacte-
rial replication, transcription, and translation [18].
Besides the establishment of a bacterial-free environment, new
biomaterials are also designed to enhance dental healing and
enamel remineralization. For example, several studies demon-
strated that the addition of Ag ions into organic or inorganic matrix
(e.g., silicate bioactive glasses) leads to more lasting ion release and
enhanced remineralization [19–22]. Such bioactive materials could
be very useful in minimal invasive techniques used in pediatric
dentistry for caries prevention. In vitro, the bioactivity of such
new biomaterials could be assessed directly by observing apatite
formation on the surface of the material after immersing the bio-
material into simulated body fluid (SBF), which imitates the ion
concentration in human plasma at physiologic conditions, 37 C
and pH 7.4 [19–21]. Alternatively, it could be assessed indirectly by
performing a microhardness test on the tooth surface surrounding
the restoration to determine the degree of demineralization.
In this model, we cultivated a microbial caries model (S. mutans
and S. sobrinus biofilm) on 4 4 bovine slabs that can be used to
assess the antibacterial effects and caries prevention potential of any
novel restorative material. The antibacterial effects can be evaluated
by examining changes in biofilm acidogenicity throughout the
incubation period and analyzing the biofilm using stereomicro-
scopy and confocal microscope system (CLSM) to assess the viabil-
ity of the bacteria and the ratio between live/dead bacteria. To
determine the bioactivity and caries prevention of the new bioma-
terial, a posttreatment Knoop surface microhardness test of the
slabs enamel around the restoration can be used as a surrogate
indicator of caries and demineralization progression.
In Vitro Caries Model 371
2 Materials
2.1 Slab Fabrication 1. Recently extracted bovine incisors (stored in 0.1% thymol).
and Initial (Baseline) 2. Water-cooled trimmer.
Microhardness Test
3. Acrylic plates and acrylic rods.
4. Low-speed saw.
5. Circular polishing machine with silicon carbide papers of
decreasing grit size.
6. High-speed handpiece and round carbide bur.
7. A stereomicroscope.
8. Restorative materials of choice (a commercially available mate-
rial to serve as control and the experimental materials of choice
to compare it to the control ones).
9. For composite resin-based materials: 37% phosphoric acid,
clear plastic strips, glass slab, and a light cure.
10. Brass holder (custom made).
11. Microhardness tester machine with a Knoop diamond
indenter.
12. 1 Phosphate-buffered saline (PBS) (pH 7.4).
13. 24-Well container.
2.2 Microbial Caries 1. Purified frozen culture of S. mutans (ATCC 700610).

Model Preparation and 2. Purified frozen culture of S. sobrinus (ATCC 27351).
Biofilm Acidogenicity
3. Brain heart infusion (BHI) agar.
Analysis
4. Autoflow CO2 water-jacketed incubator.
5. 10% sucrose solution.
6. pH electrode.
2.3 Biofilm Analysis 1. Well plates.

2. Live/Dead Bacterial Viability Kit containing two stains: the
SYTO9 stain and the propidium iodide stain.
3. Multi-chambered cover glass slides.
4. Inverted confocal microscope system.
5. Stereomicroscope.
6. Stereomicroscope camera.
7. Dental plaque disclosing agent.
3 Methods
3.1 Slab Fabrication 1. Remove the anatomical roots of the bovine incisors with a
water-cooled trimmer.
2. Mount the crowns on acrylic plates, and cut to 4 mm 4 mm
enamel/dentin slabs using a low-speed saw.
3. Ground and polish the specimens’ facial and lingual surfaces
using a circular polishing machine under water cooling (Fig. 1)
(see Note 1).
4. Attach the lingual surfaces of the specimens to acrylic rods, and
prepare a round-shaped cavity using a #4 round carbide in a
high-speed handpiece using air-water cooling.
5. Observe the specimens under stereomicroscopy; discard any
specimen that shows any irregularity at the margins of the
cavity.
6. Divide the specimens into groups: a control group where com-
mercially available restorative material will be used and experi-
mental groups for the use of the novel materials (see Note 2).
7. For composite resin-based materials: Acid etch the cavities for
15 s with 37% phosphoric acid, and then rinse and blot-
dry them.
8. Restore the cavities with the material of choice followed by a
standardized finishing/polishing procedure using different
grits of paper discs (see Note 3).
9. Glue each specimen to an acrylic rod, and mount them individ-
ually on a custom-made brass holder.
Fig. 1 4 4 bovine teeth after polishing. The specimens’ facial and lingual
surfaces were ground and polished using a circular polishing machine under
water cooling
Fig. 2 Final specimens in a 24-well container. The acrylic holders were glued to
the lids of the 24-well container, and the specimens were subsequently gas
sterilized
10. Conduct an initial baseline Knoop microhardness test at differ-

ent locations around the edges of the restoration using a
microhardness tester machine with a Knoop diamond indenter
at a 50 g load for 10 s, and record the indentations’ length (see
Note 4).
11. The specimens can be stored in phosphate-buffered saline
(PBS, pH 7.4) before incubation (see Note 5).
12. Glue the acrylic holders to the lids of 24-well container, and gas
sterilize the specimens (Fig. 2).
3.2 Microbial Caries 1. Cultivate a 0.1 mL of purified frozen culture of S. mutans

Model Preparation and UA159 and S. sobrinus (ATCC 27351) in 10 mL BHI broth
Biofilm Acidogenicity for 24 h at 37 C.
Analysis 2. Aliquot a 0.1 mL of the culture, and inoculate it to be cultured
again in fresh 10 mL BHI medium overnight at 37 C to active
culture.
3. Aliquot a 7 mL of the active culture and dilute it in 21 mL
fresh BHI.
4. Insert 2 mL of the diluted broth into each specimen well.
5. Incubate the specimens in an autoflow CO2 water-jacketed

incubator at 37 C, 10% CO2 for a proper incubation period
(see Note 6).
6. Each day, change the specimens into fresh BHI broth every
morning at 9 a.m., maintain them in 10% sucrose for 5 min
daily at 12 p.m., and then change into fresh BHI medium again
at 5 p.m., and leave it overnight, and so on.
7. Measure the pH of the overnight culture medium (old

medium) with pH electrode daily to assess acid production by
S. mutans and S. sobrinus biofilm.
3.3 Biofilm Analysis After the incubation period, wash all specimens three times in PBS.
You can evaluate the biofilm in the samples using two different
methods: (1) observe the specimens under stereomicroscope to
evaluate the total biofilm formation, and (2) observe them under
the confocal microscope using the live/dead staining to evaluate
how many of the bacteria were dead.
3.3.1 Under Confocal 1. Use at least two specimens from each group and place them in
Microscope (CLSM) well plates.
2. Stain the specimens using Live/Dead Bacterial Viability Kit
containing two stains: the SYTO9 stain and the propidium
iodide stain (see Note 7).
3. Mix equal amounts of the two dyes, and use 5 μL of the mixed
solution to stain each specimen.
4. Leave the specimens for 15 min in dark conditions, and then
put upside down in a chambered cover glass slide with PBS.
5. Observe the specimens under and inverted confocal micro-
scope system with 2-photon FLIM at 500 and 550 nm.
6. Capture two-dimensional images of specimens under a magni-
fication of 65 in several areas to analyze (Fig. 3).
7. Process and analyze the images obtained from the CLSM using
ImageJ software to determine the dead/live bacteria ratio.
3.3.2 Under 1. Stain the rest of the samples with a disclosing agent to view the
Stereomicroscope biofilm under a stereomicroscope.
2. Take pictures under a magnification of 50 with the stereomi-
croscope camera (Fig. 4) (see Note 8).
3.4 Demineralization To estimate demineralization rate, a posttreatment Knoop

Analysis (Final Knoop surface microhardness test can be used as a surrogate indicator of
Microhardness Test) caries.
1. Wipe the specimens with gauze and store in 0.1% thymol.
2. Take the posttreatment Knoop surface microhardness measure-
ments at sites immediately next to the initial Knoop ones using
the same parameters (50 g load for 10 s) (see Note 4).
3. Record the indentations length, and calculate the averages at
the different sites for data analysis.
Fig. 3 CLSM two-dimensional images taken from four different locations (a–d) of biofilm on a specimen.
Images illustrate the live/dead bacterial cell ratio in the biofilm. The SYTO9 (green) stain indicates the live
bacteria, while the propidium iodide (red) stain indicates the dead bacteria in the biofilm. Bars (a–d): 50 μm
4 Notes
1. Polish the samples by using silicon carbide papers of decreasing

grit size (#240, #400, #600, and #1200) at 100 RPM.
2. A power analysis based on paired t-test indicated a sample size
of six to be sufficient for statistical significance.
3. For composite resin-based materials: Cover the restoration
with a clear plastic strip and a glass slab to prevent the forma-
tion of the resin-rich layer after curing, and then light cure for
20 s.
4. We recommend conducting the microhardness test at four
different locations surrounding the restoration. At each loca-
tion conduct it at 50, 100, 150, and 250 μm from the edge of
the restoration, and then calculate the averages.
5. The duration of storage prior to incubation depends on the
type of material tested. For instance, when testing composite
restorations with Ag-bioglass (Ag-BGCOMP) additives, a
Fig. 4 Stereomicroscope images taken of biofilm formed on two different specimens without disclosing agent
(a, b) and with disclosing agent (c, d). Combining the data from the stereomicroscope and CLSM, images can
be used for analyzing the viability of the biofilm. Bars (a–d): 500 μm
storage period of 2 weeks prior to incubating the samples with

the bacterial culture is recommended so that silver release and
Ag-BGCOMP formation can reach its plateau.
6. We recommend incubating the samples with the bacterial cul-
ture for at least 7 days.
7. The SYTO9 stain is a green fluorescent nucleic acid stain
(emission/excitation 480–500 nm), while the propidium
iodide stain is a red fluorescent nucleic acid stain (emission/
excitation 490–635 nm). When used alone, the SYTO 9 stain
generally labels all bacteria in a population—those with intact
membranes and those with damaged membranes. In contrast,
the propidium iodide stain only penetrates the bacteria with
damaged membranes, causing a reduction in the SYTO 9 stain
fluorescence when both dyes are present.
8. We recommend taking pictures with and without the disclosing
agent.
References
1. Burke FJ, Cheung SW, Mjör IA, Wilson NH 13. Demarco FF, Collares K, Coelho-De-Souza
(1999) Reasons for the placement and replace- FH et al (2015) Anterior composite restora-
ment of restorations in vocational training tions: a systematic review on long-term survival
practices. Prim Dent Care 6:17–20 and reasons for failure. Dent Mater
2. Maupomé G, Sheiham A (1998) Criteria for 31:1214–1224
restoration replacement and restoration life- 14. Beck F, Lettner S, Graf A et al (2015) Survival
span estimates in an educational environment. of direct resin restorations in posterior teeth
J Oral Rehabil 25:896–901. https://doi.org/ within a 19-year period (1996–2015): a meta-
10.1046/j.1365-2842.1998.00328.x analysis of prospective studies. Dent Mater
3. Elderton RJ, Osman YI (1991) Preventive ver- 31:958–985
sus restorative management of dental caries. J 15. Imazato S (2003) Antibacterial properties of
Dent Assoc S Afr 46:217–221 resin composites and dentin bonding systems.
4. Drummond JL (2008) Degradation, fatigue, Dent Mater 19:449–457
and failure of resin dental composite materials. 16. Imazato S, Hua CJ, Ma S et al (2012) Anti-
J Dent Res 87:710–719. https://doi.org/10. bacterial resin monomers based on quaternary
1177/154405910808700802 ammonium and their benefits in restorative
5. Leinfelder KF (1988) Posterior composite dentistry. Jpn Dent Sci Rev 48:115–125
resins. J Am Dent Assoc 117:21E–26E 17. Kasraei S, Sami L, Hendi S et al (2014) Anti-
6. Brunthaler A, König F, Lucas T et al (2003) bacterial properties of composite resins incor-
Longevity of direct resin composite restora- porating silver and zinc oxide nanoparticles on
tions in posterior teeth: a review. Clin Oral Streptococcus mutans and lactobacillus. Restor
Investig 7:63–70. https://doi.org/10.1007/ Dent Endod 39:109–114. https://doi.org/
s00784-003-0206-7 10.5395/rde.2014.39.2.109
7. Deligeorgi V, Mjör IA, Wilson NH (2001) An 18. Moshaverinia A, Ansari S, Moshaverinia M et al
overview of reasons for the placement and (2011) Ultrasonically set novel
replacement of restorations. Prim Dent Care NVC-containing glass-ionomer cements for
8:5–11 applications in restorative dentistry. J Mater
8. Kopperud SE, Tveit AB, Gaarden T et al Sci Mater Med 22:2029–2034. https://doi.
(2012) Longevity of posterior dental restora- org/10.1007/s10856-011-4391-7
tions and reasons for failure. Eur J Oral Sci 19. Qi YP, Li N, Niu LN et al (2012) Reminerali-
120:539–548. https://doi.org/10.1111/eos. zation of artificial dentinal caries lesions by
12004 biomimetically modified mineral trioxide
9. Sakaguchi RL (2005) Review of the current aggregate. Acta Biomater 8:836–842. https://
status and challenges for dental posterior doi.org/10.1016/j.actbio.2011.10.033
restorative composites: clinical, chemistry, and 20. Sauro S, Osorio R, Watson TF, Toledano M
physical behavior considerations. Summary of (2012) Therapeutic effects of novel resin bond-
discussion from the Portland Composites Sym- ing systems containing bioactive glasses on
posium (POCOS) June 17–19, 2004, Oregon mineral-depleted areas within the bonded-
Health & Science University, Portland, Ore- dentine interface. J Mater Sci Mater Med
gon. In: Dental materials, pp 3–6 23:1521–1532. https://doi.org/10.1007/
10. Van Dijken JWV, Pallesen U (2013) A six-year s10856-012-4606-6
prospective randomized study of a nano-hybrid 21. Kokubo T (1998) Apatite formation on sur-
and a conventional hybrid resin composite in faces of ceramics, metals and polymers in body
Class II restorations. In: Dental materials, pp environment. Acta Mater 46:2519–2527.
191–198 https://doi.org/10.1016/S1359-6454(98)
11. Ferracane JL (2011) Resin composite—state of 80036-0
the art. Dent Mater 27:29–38 22. Kim HM, Miyaji F, Kokubo T, Nakamura T
12. Opdam NJM, Van De Sande FH, Bronkhorst E (1997) Apatite-forming ability of alkali-treated
et al (2014) Longevity of posterior composite Ti metal in body environment. J Ceram Soc
restorations: a systematic review and meta- Jpn 105(1218):111–116
analysis. J Dent Res 93:943–949
Chapter 34
Protocols to Study Dental Caries In Vitro: pH Cycling Models

Bennett T. Amaechi
Abstract
As laboratory models are bridges to in vivo caries studies, they must mirror clinical conditions, where
demineralization and remineralization alternate constantly (i.e., pH cycling) and are only interrupted
during the very short period of application of investigational products, such as toothpaste or mouth
rinse. In view of this, models have been developed, based on pH cycling, to study the anticaries or caries
remineralizing effects of substances. The pH cycling models have long been accepted and utilized by the
scientific community and the toothpaste industry as an appropriate alternative to animal caries testing,
particularly for ionic fluoride-based dentifrices. Several pH cycling models have been developed and
described in the literature over the years. However, in this chapter, we crudely categorize them into two
types: according to what the investigational product is tailored to achieve, i.e., prevention of caries
development (net demineralization) or remineralization of early caries (net remineralization). Thus the
models are termed “demineralization” or “remineralization” models and are described in details here
together with their disadvantages and applications.
Key words Dental caries, Demineralization model, Remineralization model, pH cycling, Dentifrices
1 Introduction
Laboratory methods are one of the key tools in caries research as

they allow for the assessment of the anticaries potential of novel
agents under highly controlled and cost-effective conditions and
thereby provide researchers with valuable information ahead of
often costly in situ and/or clinical research. However, as laboratory
models have to be seen as bridges to in vivo caries studies [1], these
laboratory models must mirror clinical conditions, where deminer-
alization and remineralization alternate constantly (i.e., pH cycling)
and are only interrupted during the very short period of application
of investigational products, such as toothpaste. Although the pH
cycling models have long been accepted and utilized by the scien-
tific community and the toothpaste industry as an appropriate
379
380 Bennett T. Amaechi
alternative to animal testing, particularly for ionic fluoride-based

dentifrices [2, 3], these models, unlike the microbial models,
exclude the cariogenic microbial biofilm that is a key factor in the
caries process. Thus the alternating demineralization and reminer-
alization cycle (caries process) occurring in a pH cycling model is
not as a result of the metabolic action of a microbial biofilm adhered
on teeth, when it is exposed to fermentable sugars, rather the feast
and famine episodes to which the biofilm is exposed to in the oral
cavity is mimicked by alternating exposure of the tooth samples to
demineralization and remineralization solutions. Specifically, a pH
cycling test is a laboratory experiment in which either sound teeth
or teeth bearing artificially induced caries lesions are exposed to
multiple dentifrice treatments, periods of demineralization, and
periods of simulated saliva interaction (remineralization) using cal-
cium phosphate solutions [2]. Demineralization periods, which is
an extended periods of acid exposure in the pH cycling protocols,
are designed to mimic the destructive effects of acid challenge to
the tooth surface and subsurface that occurs clinically, while remi-
neralization periods mimic the protective and repairing effects of
saliva [4]. A pH cycling model should show a distinct separation
between the clinically proven positive control and a fluoride-free
negative control, so that specimens receiving a fluoride-free control
treatment develop caries (if sound teeth) or progress toward cavita-
tion (if teeth containing artificial caries lesions), while specimens
receiving positive control treatment develop minimal or no caries.
It is expected that in this model, the test treatment will prevent
caries by (a) enhancing the naturally occurring remineralization
process and (b) simultaneously providing protection against the
inevitable acid attack. However, a disadvantage of the pH cycling
models is that the antibacterial effect of anticaries agents cannot be
studied with these models.
Several pH cycling models have been developed and described
in the literature over the years. However, in the current chapter, we
will crudely categorize them into two types of models, according to
what the investigational (test) product is tailored to achieve, i.e.,
prevention of caries development (net demineralization) or net
remineralization of early caries lesions. Thus we have “deminerali-
zation (caries prevention or anticaries)” model [2, 3] and “remi-
neralization” model [5]. This will go further to determine the type
of substrate to be used as specimens. A “demineralization” model
uses sound teeth to test the ability of the investigational product to
prevent caries development, while “remineralization” model uses
teeth bearing artificial caries lesions to test the ability of investiga-
tional product to cause remineralization of an artificially produced
caries lesion.
pH Cycling Models for Studying Dental Caries 381
2 Demineralization (Caries Prevention or Anticaries) Model
The demineralization model (otherwise known as caries prevention

or anticaries or lesion progression model) described in this paper is
the Featherstone pH cycling model [2–6], which was developed as
an alternative to animal caries reduction test (which is considered
the “gold standard”) required by the Food and Drug Administra-
tion (FDA) for demonstration of efficacy of “Anticaries dentifrice
drug product formulations for over-the-counter human use.” The
Featherstone pH cycling model has demonstrated excellent corre-
lation and equivalent accuracy to the currently accepted animal
caries models by the following:
(a) Demonstrating a clinically relevant fluoride dose response
similar to that shown in the animal caries model (including
1100 ppm F, 250 ppm F, and placebo).
(b) Demonstrating similar results to the animal caries model for
clinically proven dentifrice formulations relative to positive
and negative controls.
(c) Demonstrating discriminating ability in strong agreement
with the animal caries model for differentiating between a
dentifrice formulation with attenuated fluoride activity and a
USP standard.
(d) Providing a clinically relevant simulation of the effect of the
caries process, as demonstrated by orthodontic banding
studies [7].
(e) Sufficiently addressing both salivary and abrasive/anticalculus
agent interference concerns [8].
The Featherstone pH cycling model has also been shown to
correlate with the findings from human clinical trials [8]. It was also
able to discriminate between formulations that meet and those that
do not meet the anticaries monograph requirements based on the
animal caries models [9]. This model, long been accepted and
utilized by the scientific community and the toothpaste industry
as an appropriate alternative to animal testing, particularly for ionic
fluoride-based dentifrices, can also serve as a “non-inferiority” test
model.
2.1 Materials The key elements of this model are shown in Table 1. The appro-
priate clinically proven reference standard (USP NaF/silica Refer-
2.1.1 Key Elements
ence Standard Toothpaste) should serve as the positive control and
should contain the same anticaries active agent as the test product.
The negative control should be placebo toothpaste (0 ppm F). An
additional internal reference point (100 ppm F as NaF) for overall
completeness should be prepared by a 1:10 dilution (1 part prod-
uct:10 parts deionized water) of the positive control (USP NaF
Table 1
Key elements of the caries prevention model
Key element General description

Test substrate Enamel (human or bovine)
Base size (number of Sufficient to demonstrate statistically significant differences between + and –
specimens) controls and enable a sufficient power to detect desired differences
Demineralization Acid-based solution that partially demineralizes enamel, leaving enamel
solution structurally intact
Remineralization Mineral-based solution that simulates saliva and is designed to promote the
solution remineralization process
Controls Positive and negative controls over the appropriate range of products intended
to be tested
Product dilution for To simulate actual product use (toothpaste:water ratio ¼ 1:3)
testing
Treatment times 2 min, twice per day
Daily remineralization 6 h per day (between treatment periods)
exposure
Daily remineralization 16 h per day (overnight)
exposure
Specimen analysis Quantitative assessment of mineral profile from the enamel surface to a depth
of constant mineral (i.e., sound mineral) beneath a formed lesion
Statistical power Minimum requirement: 80% power to separate a 20% difference defined
Reference Standard) with distilled water. All test products (includ-

ing positive and negative controls and the internal reference point)
are provided to the laboratories in blank, blinded tubes.
2.2 Methods Sound human or bovine teeth are collected and cleansed of soft
tissue debris, brushed with pumice slurry using an electric tooth-
2.2.1 Sample
brush, and then examined by transillumination. Teeth without
Preparation
cracks, hypoplasia, white spot lesions, and other malformations
are selected. The roots of each tooth are cut off. Whole tooth or
tooth slabs can be used. The sample analysis method also deter-
mines the specimen preparation. If surface microhardness analysis
will be used, then tooth slabs is used, and both the top and bottom
of the slab are polished to achieve flat and planoparallel surfaces
required for surface microhardness (SMH) measurement. The top
should also be polished to high luster required for analysis. All
surfaces of each tooth or slab is painted with two coats of acid-
resistant nail varnish, except for a window of exposed enamel to be
created on the buccal surface for development of caries.
2.2.2 Sample Size Each of the selected teeth is randomly assigned to the experimental
treatment groups, with a minimum of ten teeth/group. A number
of specimens must be sufficient to support statistical separation of
positive and negative controls and enable a sufficient power to
detect desired differences. The specimens should be mounted in a
manner to facilitate handling.
2.2.3 Measurement of The baseline surface microhardness (SMHb) of the tooth blocks are
Baseline Surface measured on each selected tooth block using a Knoop diamond
Microhardness (SMH) indenter (Tukon 2100; Wilson-Instron, Norwood, MA, USA),
with a load of 50 g applied for 5 s. The measurement is made at
the exposed enamel window (2 mm diameter). Three indentations
are made at the middle, upper, and lower ends of the enamel surface
(preserving a reasonable sound area between the indentations), and
the Knoop numbers are calculated and averaged for each block. The
distance between the indentations is measured.
2.2.4 Solution Standardized demineralization and remineralization solutions [10]

Preparation are prepared according to the compositions shown in Table 2. The
demineralizing solution served as an acid challenge similar to those
generated by plaque acids in the mouth. The demineralization
solution should provide for a partial demineralization of the
enamel, leaving the enamel softened, yet structurally intact, which
is morphologically similar to the human caries condition. Fresh
demineralization solution is changed twice weekly. A mineral solu-
tion is used as the remineralization medium in all treatment regi-
mens. The remineralization solution should be demonstrated to be
capable of enhancing the natural process of remineralization. Fresh
mineral mix is changed three times per week. Solutions are stored in
sealed containers at room temperature throughout each of the
experiments.
2.2.5 Test Products If the test products are toothpaste, freshly prepared slurries of both
Preparation test and control toothpaste are made with deionized distilled water,
pooled human saliva, or artificial saliva. A dilution of one part
toothpaste to three parts diluent, thoroughly mixed for 4 min,
using a laboratory stand mixer until homogenous; 4.0 mL per
tooth of 1:3 with diluent is recommended, as this represents the
anticipated level of dilution that occurs during routine use of
toothpaste products.
2.2.6 Treatment The cyclic treatment regimen for each day is shown in Table 3. It
Regimen consist of two 2-min toothpaste treatment periods, one 6-h acid
challenge, and then storage in remineralizing solution for the rest
of the time, including night. Specimens are treated with freshly
prepared slurries of toothpaste two times per day. For treatment,
the demineralization and remineralization solutions are magneti-
cally stirred at a speed of 350 rpm, while the toothpaste slurry is
384
Table 2
Summary of the study protocol, including the composition of demineralizing and remineralizing solution
Base Daily Daily

size Test demin remin
Test per Demineralization Remineralization toothpaste time time Evaluation of
substrate group solutiona solutionb dilution ratio Diluents Treatment (h) (h) substrate
Bennett T. Amaechi
Human 10 Calcium: Calcium: 1:3 Deionized 2 min 2/day for 6 16 Cross-sectional

enamel 2.0 mmol/L 1.5 mmol/L toothpaste: distilled 14 days (human microhardness
(Intact 0.4723 g/L 0.3542 g/L diluent water enamel) and 9 days
surface) Ca(NO3)2·4H2O Ca(NO3)2·4H2O (bovine enamel)
Mwt ¼ 236.16 Mwt ¼ 236.16
Phosphate: Phosphate: Pooled Surface
2.0 mmol/L 0.9 mmol/L human microhardness
0.2722 g/L 0.1225 g/L saliva
KH2PO4 KH2PO4
Mwt ¼ 136.09 Mwt ¼ 136.09
Acetic acid: KCI: 130 mmol/L Artificial Transverse
75.0 mmol/L 9.6915 g/L saliva microhardness
4.5083 g/L KCl
CH3COOH Mwt ¼ 74.55
Mwt ¼ 60.05
pH 4.5 NaCacodylate:
20 mmol/L
4.28 g/L
NaC2H6AsO2·3H2O
Mwt ¼ 214
pH 7.0
a
Adjusted to appropriate pH with 50% NaOH after all ingredients were dissolved completely
b
Adjusted to pH 7.0 with concentrated HCI
Table 3
pH cycling treatment sequence for the experiment
Time Treatment
Day 1 is all-day storage in remineralization solution. Then, subsequent days’ treatments will be as follows
2 min (starts 8:00 a.m.) Toothpaste treatment
Approximately 1 h to complete all groups
Rinse with deionized distilled water
6 h (9:00 a.m.–3:00 p.m.) Acid challenge (demineralization)
2 min (starts 3:00 p.m.) Toothpaste treatment
Approximately 1 h to complete all groups
16 h (from 4:00 p.m. till 8:00 a.m. next day) Storage in remineralization solution
Repeat for 13 additional days (human) and 8 additional days (bovine)
static. Specimens can be treated individually or collectively as treat-

ment groups. Whatever the case, the volume of solution should be
40 mL of demineralizing solution per specimen and 20 mL of
remineralizing solution per specimen. All treatments are carried
out in an incubator at 37 C. The pH of each medium is measured
once daily before treatment. The regimen starts with 2-min treat-
ment with toothpaste slurry, after which the specimens are rinsed
with running deionized water and dried with paper towel before
immersion into the demineralizing solution for 6 h at 37 C. After
demineralization, specimens are rinsed with deionized water and
treated again with toothpaste slurries. Specimens are then rinsed
with deionized water and placed into remineralization solution for
16 h overnight. This daily pH cycling regimen is repeated for a total
of 14 days with the human enamel and for a total of 9 treatment
days with the bovine enamel.
2.2.7 Posttreatment On termination of the experiment, the teeth will be harvested and
Evaluation of Substrate processed for demineralization assessment using either cross-sectional
microhardness through the depth of the lesion and into the underly-
ing sound enamel or surface microhardness. An alternative to cross-
sectional microhardness is the use of quantitative transverse microra-
diography (TMR), which has been demonstrated to provide data that
correlates with cross-sectional microhardness [11–13].
Posttreatment Surface The SMHT measurement should be performed as described above

Microhardness by three indentations on the free (un-indented) surface of the block
Measurement (SMHT) and the average value calculated for each block. At this point the
pretest (SMHb) and posttest (SMHT) surface microhardness value
of the lesions is available for data analysis.
Cross-Sectional Following SMH measurement, the tooth slab (tooth) is sectioned

Microhardness into two halves perpendicularly through the center of the exposed
Measurement (CSMH) window. One of the cut halves from each specimen, chosen at
random, is mounted and used for analyses. These specimens from
each test group are mounted in resin to cover all surfaces except the
cut faces. The cut faces left exposed are flattened and then serially
polished using abrasives from high (~10 μm) down to low (~l μm)
in order to provide a high luster for microhardness analyses. Cross-
sectional microhardness is performed as previously described [11],
using a Knoop diamond from the specimen surface and down
through the depth of the lesion and into the underlying sound
enamel.
Transverse From the remaining half of the slab or tooth, a tooth slice
Microradiography and (150 μm thick) is cut perpendicularly to the exposed window in
Image Analysis each tooth specimen using a water-cooled saw. Each slice is
polished to 100 μm thickness. Both sides of the slice are polished
using Adhesive Back 6 μm lapping film in a MultiPrep™ Precision
Polishing machine (Allied High Tech, USA) to achieve planopar-
allel surfaces as well as to reduce the thickness of the slice to 100 μm
(the appropriate thickness for TMR). Then the slices are microra-
diographed on type lA high-resolution glass X-ray plates (Micro-
chrome Technology, CA, USA) using a Phillips X-ray generator
system (Panalytical, Amsterdam) setup for this purpose. The plates
are exposed for 10 min at an anode voltage of 20 kV and a tube
current of 10 mA and then processed. Processing consisted of a
5 min development in Kodak HR developer and 15 min fixation in
Kodak Rapid Fixer before a final 30 min wash period. After drying,
the microradiographs are subjected to visualization with a Leica
DMR optical microscope linked via a Sony model XC-75 CE
CCTV camera to a computer housing the image analysis program
(TMR2006 version 3.0.0.6, Inspektor Research, Amsterdam). The
enhanced images of the microradiographs are analyzed under stan-
dard conditions of light intensity and magnification and processed,
along with data from the image of the step wedge, by the TMR
program. The computer program calculates the parameter of
integrated mineral loss (vol%.μm) and the lesion depth (μm)
based on the work described by De Josselin de Jong et al. [14].
The integrated mineral loss was defined as the difference in volume
percent of mineral between sound and demineralized tissue
integrated over the lesion depth. The lesion depth was assessed as
the distance from the measured sound enamel surface to the loca-
tion in the lesion at which the mineral content is greater than 95%
of the mineral content in sound enamel. By this method, the
parameters of integrated mineral loss (Δz, vol%.μm) and lesion
depth (LD, μm) are quantified for each caries lesion.
2.2.8 Data Management (1) For SMH data, the mean values of the SMHb and SMHT will
and Statistical Analysis be calculated for each treatment group and be compared using
paired t-test to determine if there is any significant change
(demineralization) in SMH (intragroup comparison). To con-
duct intergroup comparisons between the toothpaste groups,
percentage change in SMH (%ΔSMH), calculated relative to
the baseline (SMHb), will be determined for each test product
(percentage change is used for comparison in order to make
provision for the fact that the tooth blocks in all groups came
from different teeth and as such the SMH for the blocks may
differ at baseline). This is calculated thus:
SMHb SMHT
% change in SMHð%ΔSMHÞ ¼ 100
SMHb
Using the mean values of the %ΔSMH, the toothpaste
groups are compared among themselves according to the
recommended statistical tests and the groups being
compared.
(2) For CSMH data, the Knoop hardness is measured at different
depths into the tooth tissue in two lanes, the demineralized
(lesion) lane on the exposed enamel window and the sound
enamel lane on the area protected by acid-resistant nail var-
nish. After assessing and capturing all the Knoop indentations
lengths from both sound and lesion areas on each depth, the
percent difference between the CSMH values of the lesion
and sound lanes at each depth is calculated. An example
calculation for the first indentation made in each lane of a
given specimen would appear as such:
CSMHLesion CSMHSound
% CSMH Diff Depth ¼ 100
CSMHSound
The use of this equation produces a negative percent
difference if the value in the lesion region is lower than the
value in the sound region, thereby indicating tooth deminer-
alization. The Knoop hardness for the sound and lesion mea-
surement lanes can be compared directly.
Alternately, if a single value is desired in addition to %
CSMH Diff at each of the depths, then the average percent
difference in CSMH for each lane (i.e., the lane made under
the sound enamel surface and the lane made under the demi-
neralized surface) can be calculated and using the similar form
of the equation above can be expressed as:
% CSMHDiff AverageLesion % CSMHDiff AverageSound
Average % CSMH Diff ¼ 100
% CSMHDiff AverageSound
In this equation, the % CSM DiffAverageLesion
(or AverageSound) is determined from each of the % CSM
DiffDepth calculated at each depth.
(3) For TMR data, the TMR process will yield the following
information:
(a) The mineral loss and lesion depth of any lesions
(b) The TMR images of the lesions
Using the TMR images, the extent of demineralization pro-
duced within the internal structure of each specimen in each
group will be examined and described. The TMR data for the
independent experiments should be analyzed according to the
recommended statistical tests and the groups being
compared.
3 Remineralization (Net Remineralization) Model
As stated above, “remineralization” model is a pH cycling model

used when the investigational (Test) product is tailored to promote
remineralization of an early-stage caries lesion by enhancing the
naturally occurring remineralization process of saliva. Hence, a
remineralization model uses teeth or tooth slabs bearing artificially
produced caries lesions, expected to be remineralized by the test
product. A typical “net remineralization” model was described and
used by Lippert et al. [15] and was also validated by measurement
of fluoride uptake into the caries lesions following remineralization
[15]. The model has been validated for artificial early caries created
in human and bovine enamel [16] as well as in fluorotic and sound
human teeth [17].
3.1 Materials Sound human or bovine teeth can be used. Teeth are collected and
sterilized in accordance with the university procedure for steriliza-
3.1.1 Specimen
tion of teeth used for studies. Following sterilization, the teeth are
Preparation
cleansed of soft tissue debris, brushed with pumice slurry, prefera-
bly using an electric toothbrush, and then examined by transillumi-
nation. Teeth without cracks, hypoplasia, white spot lesions, and
other malformations are selected. The teeth can be stored in dis-
tilled deionized water saturated with thymol during the sample
preparation process. Using a water-cooled cutter, tooth blocks
(approximately 3 mm length 3 mm width 1.5 mm thick) are
produced from the buccal and lingual surfaces of each tooth. Using
a Rotopol 31/Rotoforce polishing unit, specimens are grounded
and polished to create flat planoparallel dentin and enamel surfaces
required for surface microhardness (SMH) measurement. Then the
enamel surface is serially polished to luster using adhesive back
lapping film (last film 30 μm) in a MultiPrep™ Precision Polishing
machine. Following this, all surfaces of each block are painted with
two coats of acid-resistant nail varnish, except a window on the
enamel surface. Prepared specimens are stored at 100% relative
humidity at 4 C until use. All specimens are prepared by the
same trained technicians using standard operating procedures.
3.1.2 Sample Size Each specimen is randomly assigned to the experimental treatment
groups. A number of specimens in each treatment group must be
sufficient to support statistical separation of positive and negative
controls and enable a sufficient power to detect desired differences.
The specimens should be mounted in a manner to facilitate handling.
3.2 Methods Early caries-like lesions are created on the exposed window on each
tooth block by subjecting the blocks to 7 days (4 days for bovine)
3.2.1 Caries Lesion
demineralization in an acidified gel system. The gel is prepared by
Formation
adding 0.10 M sodium hydroxide to 0.10 M lactic acid to give a
final pH value of 4.5. To this solution, 6% w/v hydroxyethyl cellu-
lose is added while vigorously stirring. The final consistency of gel
achieved should have a viscosity in the region of 100 cP. Deminer-
alization periods were chosen based on prior experience and to
create lesions with comparable data. Following lesion formation,
the nail varnish will be carefully and totally removed with acetone.
3.2.2 Lesion Baseline Decide your measurement method(s). Two common standard
Measurement measurements are (a) microhardness (surface [SMH] or cross-
sectional [CSMH]) and (b) transverse microradiography (TMR).
One or a combination of two or even the entire three measurement
methods can be performed in a single study.
(a) If SMH analysis is to be used, determine the initial SMH of the
demineralized specimens using a Vickers microhardness
indenter at a load of 200 g for 15 s [5]. Determine average
specimen surface microhardness (VHNbase) from four inden-
tations on the surface of each specimen. Exclude slabs with
very low or very high surface hardness values (e.g., select slabs
within a range of 10–20% above or below the average hardness
value of all slabs) and those with a great variability among the
indentation values (e.g., exclude slabs with a coefficient of
variation of the 3–5 indentations greater than 10%). Speci-
mens are assigned to groups following a stratified randomiza-
tion procedure, based on their VHNbase.
(b) If CSMH analysis is to be used, section the tooth slab into
two, cutting through the center of the lesion. Then use one
half to determine the initial CSMH of the demineralized speci-
mens as described previously [11], using a Knoop diamond
from the specimen surface and down through the depth of the
lesion and into the underlying sound enamel.
(c) An alternative to cross-sectional microhardness is the use of
quantitative transverse microradiography (TMR), which has
been demonstrated to provide data that correlates with cross-
sectional microhardness [11–13]. If TMR is to be used,
instead of using the half slab for CSMH, a slide should be
cut from that half and processed for TMR analysis as described
under Subheading “Transverse Microradiography and Image
Analysis”.
3.2.3 Test Products If the investigational (test) products are toothpaste, freshly
Preparation prepared slurries of both test and control toothpaste are made
with either deionized distilled water, pooled human saliva, or artifi-
cial saliva, depending on what is to be used as the remineralizing
solution. A dilution of one part toothpaste (9 g) to three parts
remineralizing solution (27 mL) is thoroughly mixed for 4 min in a
beaker using a magnetic stirrer or a laboratory stand mixer until
homogenous; 4.0 mL per tooth of 1:3 with diluent is recom-
mended, as this represents the anticipated level of dilution that
occurs during routine use of toothpaste products. Fresh slurry
should be prepared for each group just prior to each treatment.
3.2.4 Solutions Standardized remineralization and demineralization solutions are

Preparation prepared. The remineralization solution should be demonstrated to
be capable of causing remineralization of an early caries lesion.
Preferably, a 1:1 mixture of human saliva and artificial saliva
(2.20 g/L gastric mucin, 1.45 mM CaCl2·2H2O, 5.42 mM
KH2PO4, 6.50 mM NaCI, 14.94 mM KCI) can be used as the
remineralization medium. Saliva can be previously collected on ice
from at least five healthy volunteers chewing paraffin wax, after at
least 2 h of fasting, pooled and then centrifuged at 3800 g for
10 min at 4 C. Collect supernatants and filter using a 0.2 μm filter.
Filtered saliva can be stored on ice until use. Fresh saliva mixture
should be used each day (changed during the acid challenge
period).
The demineralizing solution served as an acid challenge
similar to those generated by plaque acids in the mouth. The
demineralization solution should preferably consist of 2.0 mMol/
L Ca(NO3)2.4H2O, 2.0 mMol/L KH2PO4, 75.0 mMol/L
CH3COOH with pH adjusted to 4.5 using KOH.
3.2.5 pH Cycling The cyclic treatment regimen for each day is shown in Table 4. The
Regimen daily cyclic treatment regimen consists of a 4-h acid challenge in the
demineralization solution and four 2-min dentifrice slurry treat-
ment periods with specimens stored in a 1:1 mixture of pooled
human/artificial saliva all other times. Dentifrice slurry and saliva
treatments are stirred at 350 rpm, whereas the demineralization
treatment is not. After each treatment, the specimens are rinsed
briefly under running deionized water. All specimens are then
placed back into the saliva mixture. The experimental phase is
conducted at room temperature. The treatment schedule as out-
lined in Table 4 is followed daily over a period of 20 days.
3.2.6 Posttreatment (a) The mean VHNpost of each specimen could be determined as
Evaluation of Substrate described under Subheadings “Posttreatment Surface Micro-
hardness Measurement (SMHT)” and 2.2.8, step 1 from four
indentations on the surface of each specimen, next to the
Table 4
Treatment schedule for the pH cycling phase
The following treatment schedule is not given on the first day rather the specimens are stored in
saliva with constant gentle rotation to allow an artificial pellicle-like layer to form
Time pH cycling phase

2 min (8:00–8:02 a.m.) Test/placebo treatment
1 h (8:02–9:00 a.m.) Artificial saliva (remineralization solution)
2 min (9:00–9:02 a.m.) Test/placebo treatment
1 h (9:02–10:00 a.m.) Artificial saliva
4 h (10:00 a.m.–2:00 p.m.) Acid challenge (demineralization)
1 h (2:00–3:00 p.m) Artificial saliva
2 min (3:00–3:02 p.m.) Test/placebo treatment
1 h (3:02–4:00 p.m.) Artificial saliva
2 min (4:00–4:02 p.m.) Test/placebo treatment
Overnight (4:02 p.m.–8:00 a.m.) Artificial saliva
baseline indentations. The change in VHN vs. lesion baseline

was calculated as follows (*REM > 0 indicates the ability of a
treatment to enhance remineralization after 20 days of
treatments):
REM∗ ¼ VHNpost VHNbase

(b) If TMR analysis is desired, a tooth slice (150 μm thick) is cut
from the lesion and processed for TMR analysis as described
under Subheadings “Transverse Microradiography and Image
Analysis” and 2.2.8, step 3.
(c) If CSMH is desired, the tooth slab should be cut into two
halves, and one of the two halves should be processed for
CSMH data as described under Subheadings “Cross-Sectional
Microhardness Measurement (CSMH)” and 2.2.8, step 2.
3.2.7 Statistical Analysis The data for the independent experiments should be analyzed
according to the recommended statistical tests and the groups
being compared.
References
1. White DJ (1992) The comparative sensitivity 2. Featherstone JD, Stookey GK, Kaminski MA,
of intra-oral, in vitro and animal models in the Faller RV (2011) Recommendation for a
profile evaluation of topical fluorides. J Dent non-animal alternative to rat caries testing.
Res 71:884–894 Am J Dent 24(5):289–294
3. Stookey GK, Featherstone JD, Rapozo- 10. ten Cate JM, Timmer K, Shariati M, Feather-
Hilo M, Schemehorn BR, Williams RA, Baker stone JD (1988) Effect of timing of fluoride
RA, Barker ML, Kaminski MA, McQueen CM, treatment on enamel de- and remineralization
Amburgey JS, Casey K, Faller RV (2011) The in vitro: a pH cycling study. Caries Res
Featherstone laboratory pH cycling model: a 22:20–26
prospective, multi-site validation exercise. Am J 11. Featherstone JDB, ten Cate JM, Arends J,
Dent 24(5):322–328 Shariati M (1983) Comparison of artificial
4. Faller RV, Pfarrer AM, Eversole SL, Cox ER, caries-like lesions by quantitative microradiog-
Landrigan WF, Wang Q (1997) The compara- raphy and microhardness profile. Caries Res
tive anticaries efficacy of Crest toothpaste rela- 17:385–391
tive to some marketed Chinese toothpastes. 12. Arends J, ten Bosch JJ (1992) Demineraliza-
Results of in vitro pH cycling testing. Int tion and remineralization evaluation techni-
Dent J 47:313–320 ques. J Dent Res 71:924–928
5. White DJ (1987) Reactivity of fluoride denti- 13. Kielbassa AM, Wrbas KT, Schulte-Meriting J,
frices with artificial caries. 1. Effects on early Hellwig E (1999) Correlation of transversal
lesions—F-uptake, surface hardening and microradiography and microhardness on in
remineralization. Caries Res 21:126–140 situ induced demineralization in irradiated
6. Featherstone IDB, Glena R, Shariati M, Shields and nonirridiated human dental enamel. Arch
CP (1990) Dependence of in vitro deminerali- Oral Biol 44:243–251
zation of apatite and remineralization of dental 14. de Josselin de Jong E, Tenbosch JJ, Noordman
enamel on fluoride concentration. J Dent Res J (1987a) Optimised microcomputer guided
69:620–625 quantitative microradiography on dental
7. O’Reilly MM, Featherstone IDB (1987) mineralised tissue slices. Phys Med Biol
Demineralization and reminerali-zation 32:887–899
around orthodontic appliances: an in vivo 15. Lippert F, Newby EE, Lynch RJ, Chauhan VK,
study. Am J Orthod Dentofac Orthop Schemehorn BR (2009) Laboratory assess-
92:33–40 ment of the anticaries potential of a new denti-
8. Lu KH, Yen DJC, Zacherl WA, Ruhlman CD, frice. J Clin Dent 20(2):45–49
Sturzenberger OP, Lehnhoff RW (1985) The 16. Lippert F, Hara AT (2012) Fluoride dose-
effect on dental caries of a fluoride dentifrice response of human and bovine enamel caries
containing an anti-calculus agent. J Dent Child lesions under remineralizing conditions. Am J
52:449–451 Dent 25(4):205–209
9. Landrigan WF, Eversole SL, Best JM, Faller RV 17. Alhawij H, Lippert F, Martinez-Mier EA
(1998) Animal caries efficacy of conventional (2015) Relative fluoride response of caries
and ‘remineralizing’ toothpastes. J Dent Res lesions created in fluorotic and sound teeth
77(Sp Iss B):1693 studied under remineralizing conditions. J
Dent 43(1):103–109
Chapter 35
In Vivo Rodent Models for Studying Dental Caries

and Pulp Disease
June Hsiao, Yuanyuan Wang, Li Zheng, Ruirui Liu, Raed Said,
Lubomir Hadjiyski, Heekon Cha, Tatiana Botero, Xanthippi Chatzistavrou,
Qing Dong, Silvana Papagerakis, and Petros Papagerakis
Abstract
Dental caries is an infectious oral disease caused primarily by complex interactions of cariogenic oral flora
(biofilm) with dietary carbohydrates on the tooth surface over time. Streptococcus mutans and Streptococcus
sobrinus (S. mutans and S. sobrinus) are the most prevalent cariogenic species within the oral biofilm and
considered the main etiological agents of caries. Pulp exposure and infection can be caused by trauma,
carious lesion, and mechanical reasons. Pulp response to these exposures depends on the state of the pulp as
well as the potential bacterial contamination of pulp tissue. Herein, we describe the process of using two
in vivo rodent models to study the progression of dental caries and pulp disease: a nutritional microbial
model and a pulp disease induction model. The progression of the carious lesion and pulpal infections in
both models was assessed by micro-CT imaging and histomorphometric analysis. Moreover, the pulp
disease induction models can be used to compare and assess the antibacterial and reparative properties of
the different pulp capping materials.
Key words Caries, Pulp, Pulp capping, Streptococcus mutans, Streptococcus sobrinus
1 Introduction
Dental caries is one of the most prevalent multifactorial diseases

that results from the interaction of cariogenic oral flora with fer-
mentable dietary carbohydrates on the tooth surface over time [1].
The main reason for enamel demineralization is dental plaque
(biofilm) which contains a complex structure of microorganisms
that can form naturally and bind on the tooth structure and around
restoration margins [2, 3]. The bacterial composition of the dental
plaque is relatively stable despite minor changes in the oral environ-
ment [4]. To date, more than 750 different species of bacteria have
June Hsiao and Yuanyuan Wang contributed equally to this work
393
394 June Hsiao et al.
been found in mature dental caries. Streptococcus mutans are not the
predominant microorganism in dental plaque, but they are an
essential component in the process of caries formation [5]. In
addition, Streptococcus sobrinus (S. sobrinus) and Lactobacillus spp.
are also implicated in the pathogenesis of dental caries. The binding
sites for cariogenic bacteria result from the complex interactions
between sucrose and the cariogenic bacteria exoenzymes [5]. This
binding will initiate the metabolic activity of microorganisms lead-
ing to acidification of the environment and eventually dissolution of
enamel [5]. Demineralization of the enamel can be reversed and
balanced by a remineralization process which is controlled by sev-
eral factors such as salivary flow, salivary components (fluoride,
calcium, and phosphate), antibacterial materials (fluoride, chloro-
hexidine, and xylitol), and good oral hygiene [6, 7]. Repeated
dissolution events of enamel can shift the balance toward deminer-
alization of the tooth structure which eventually results in the
formation of cavities [8].
Dental pulp pathology is primarily the infection of the dental
root canal system. It is also the major etiologic agent of apical
periodontitis [9]. There are several routes through which micro-
organisms can reach the dental pulp. Pulp exposure can be caused
by trauma, carious lesion, and mechanical reasons. Pulp response to
these exposures depends on the state of the pulp and its subsequent
reaction as well as a potential bacterial contamination of pulp tissue
[10]. Although it has been suggested that the bacteria from deep
periodontal pockets might reach the root canals of these teeth
through severed blood vessels of the periodontium, the exposure
of the dental pulp to the oral cavity is still the most important route
of endodontic infection [11]. Direct pulp capping of exposed, vital
painless pulps aims to maintain pulpal health, promote pulp heal-
ing, and prevent further pulp injury and subsequent treatment
[12]. A number of materials have been suggested for use in direct
pulp capping, such as calcium hydroxide (CH) and mineral trioxide
aggregate (MTA).
Bacterial profiles of the endodontic microbiota vary from indi-
vidual to individual but are generally dominated by gram-negative
organisms [13]. This indicates that primary pulpal infection has
heterogeneous etiology, where no single species can be considered
as the main endodontic pathogen, and multiple bacterial combina-
tions play a role in disease [14]. In general, during the early stage of
pulpal infection facultative anaerobic bacteria dominate consuming
most of the available oxygen, which progressively favors the growth
of obligate anaerobic bacteria [15–17]. In the later stage of infec-
tion, low oxygen tension further suppresses facultative microorgan-
isms in the dental pulp and favors anaerobic bacteria growth.
In Vivo Caries Model 395
Lipopolysaccharides (LPS) form an integral part of the outer

layer of gram-negative cell walls. They are released during disinte-
gration of bacteria after death and are also shed in small quantities
during multiplication and growth. As gram-negative organisms
dominate the endodontic flora, it is not surprising that they may
multiply and die in the root canal, thereby releasing LPS that can
egress through the apical foramen into the periapical area to initiate
and sustain apical periodontitis [17]. LPS acts as the prototypical
endotoxin as it promotes the secretion of pro-inflammatory cyto-
kines, nitric oxide, and eicosanoids [18]. In the initial stages of pulp
infections, these inflammatory mediators are critical drivers of the
repair process and stimulate reparative dentin formation by odon-
toblasts and the differentiation of progenitor cells into a repair
phenotype [19]. In addition, odontoblasts have also been shown
to induce neutrophil migration via IL-8 secretion in response to
LPS stimulation [20]. However, if the mediators persist longer,
inflammation becomes sustained in the pulp, creating cell toxicity
and tissue disrupting effects, ultimately leading to tissue
necrosis [21].
Herein, we describe the process of utilizing two in vivo rodent
models to study the progression of dental caries and pulp disease
over time: (1) a nutritional microbial model where a bacterial caries
biofilm was cultivated and inoculated into the rats’ teeth followed
by administration of high-sugar diet over an extended period of
time (6 weeks) to mimic the human caries etiology and progression
(2) and a pulp disease induction model where a pulp exposure site
was created and then immersed with LPS endotoxin to induce
pulpal infection. The progression of the carious lesion and pulpal
infections is assessed by micro-CT imaging and histomorphometric
analysis. Moreover, the pulp disease induction models can be also
used to compare and assess the antibacterial and reparative proper-
ties of the different pulp capping materials.
2 Materials
2.1 Nutritional Prepare all solutions using ultrapure water (Milli-Q water). Use
Microbial Bacterial sterilized ultrapure water for drinking water for rats.
Model
1. 21 days (just weaned) pathogen-free Sprague–Dawley (SD) rats
(see Note 1).
2. Purified frozen culture of S. mutans (ATCC 700610).
3. Purified frozen culture of S. sobrinus (ATCC 27351).
4. Brain heart infusion (BHI) agar.
5. Powdered high-sugar diet 2000 (Harlan Teklad lab,
Indianapolis, IN).
6. 5% sucrose for drinking (see Note 2).
Fig. 1 Dental clinic microbrush for applying the bacterial samples on the rats’
teeth
7. Dental clinic use microbrush (Fig. 1).

8. Ketamine and xylazine anesthetics.
9. 4% paraformaldehyde (PFA) (pH 7.4) (see Note 3).
10. Micro-CT scanner with MicroView software.
11. 70% ethanol solution (ETOH).
12. Decalcification solution; 0.5 M ethylenediaminetetraacetic acid
(EDTA) (pH 8) (see Note 4).
13. Hematoxylin and eosin stain solution.
2.2 Pulp Disease 1. Sprague–Dawley adult rats.

Induction Models 2. Ketamine/xylazine anesthetics.
3. Rat stabilizer with retraction hooks.
4. Dissection microscope.
5. 5 μg/mL lipopolysaccharide (LPS) from Bacteroidetes (Bac-
teroids + Porphylomonas) (Sigma Aldrich, St. Louis, MO,
USA).
6. Bonding agent applicator.
7. Normal saline.
8. Pulp capping materials of choice.
9. Micro-CT scanner with MicroView software,
10. 70% ethanol solution (ETOH).
11. Decalcification solution; 0.5 M ethylenediaminetetraacetic acid
(EDTA) (pH 8) (see Note 4).
12. Hematoxylin and eosin stain solution.
3 Methods
3.1 Nutritional 1. Just weaned (21 days) Sprague–Dawley (SD) rats should be
Microbial Bacterial used to avoid bacteria contamination. Feed rats with Diet 2000
Model (powdered high-sugar diet) by using powder feeder jar. Give
5% sucrose for drinking. Diet and drinking are provided ad
libitum.
2. Prepare 2.5 109 CFU of Streptococcus mutans and Streptococ-
cus sobrinus in BHI broth for each rat. Centrifuge bacteria at
2000 rpm for 3 min. Discard culture medium. Resuspend
bacteria with 1 mL PBS.
3. Anesthetize rats with ketamine and xylazine following manu-
facture’s instruction (see Note 5).
4. Inoculate Streptococcus mutans and Streptococcus sobrinus on rat
teeth using a microbrush. Reapply the bacteria on teeth every
week for 6 weeks.
5. Sacrifice rat and dissert heads. Fix samples with 4% PFA for
overnight. Change samples to 70% EOTH and wait for micro-
CT scanning for caries analysis.
6. Scan the samples with micro-CT in 18 micrometer resolution
(see Note 6). Use MicroView software to analysis micro-CT
images and find out the locations of caries on rat teeth (Fig. 2).
7. Dissect maxilla and mandible. Decalcify samples with 0.5 M
EDTA for 3 weeks.
Fig. 2 (a, b) Micro-CT analysis of interproximal caries in S. mutans- and S. sobrinus-infected rats. Interproxi-
mal caries were detected and counted by micro-CT (arrows)
Fig. 3 (a) The rat’s stabilizer used in this protocol. (b) Two hooks were attached to the maxillary and
mandibular incisors of the rat to maintain sufficient access during pulp exposure and subsequent treatments
8. Embed in paraffin and section at 5 μm.

9. Perform hematoxylin and eosin staining to visualize the caries
damage in dentin.
3.2 Pulp Disease Induce pulp disease after 1 week of adjustment of animals to the
Induction Models environment; perform all procedures with the aid of a microscope
at 10 magnification.
1. Anesthetize the animals with ketamine/xylazine according to
the manufacturer instructions (see Note 5).
2. Stabilize the rats on an operation bed (Fig. 3).
3. Attach two hooks to the maxillary and mandibular incisors of
the rat to achieve a sufficient mouth opening (Fig. 3).
4. Isolate the right and left maxillary first molars.
5. Create a pulp exposure with a size of ¼ round bur (0.5 mm) on
the maxillary first molars mesial-occlusal surfaces (Fig. 4).
6. Keep the cavity open for 2 min, and then apply the LPS
solution into the cavity preparation with a bonding applicator.
7. Keep the cavities unfilled for 24 h in order to induce dental
pulp inflammation.
8. After the 24 h have passed, anesthetize the animals again with
ketamine/xylazine and stabilize them in the same way.
9. Rinse all the tested teeth with normal saline for 5 s and dry with
air, and then apply the capping material of choice onto the
exposed cavity (Fig. 5) (see Note 7).
10. Euthanize the rats 8 weeks after treatment with 13–30% of
carbon dioxide, dissect the maxillae, and fix them with 70%
ethanol.
Fig. 4 (a) Illustration of the exposure sites in this animal model. The red arrows point to the pulp exposure
sites. (b, c) Clinical photos of pulp exposure (arrows): (b) before pulp exposure and (c) after pulp exposure
(notice the bleeding)
11. Scan over the entire length of the sample using a micro-CT
system at 18 μm resolution (Fig. 6) (see Note 6).
12. After scanning with the micro-CT, fix the specimens in 70%
ethanol.
13. Decalcify samples with 0.5 M EDTA.
14. Embed in paraffin and section at 5 μm.
15. Perform hematoxylin and eosin staining to visualize the caries
damage in dentin (Fig. 7) (see Notes 8–10).
4 Notes
1. Just weaned (21 days) Sprague–Dawley (SD) rats should be

used to avoid bacteria contamination.
2. Sterilization is necessary for 5% sucrose solution.
3. To make 20% PFA, add 200 g of paraformaldehyde powder and
100 mL of 10 PBS to 700 mL of water, heat the solution to
about 55–60 C, and adjust pH by slowly adding 1 N NaOH.
The powder will slowly dissolve when pH increased, and the
solution will be clear. Remeasure the pH and adjust to 7.4. The
solution can be filtered, aliquoted, and frozen in freezer.
Fig. 5 (a) Illustration of the sequences of preparation for pulp exposure and restoration. The cavities remained
unfilled for 24 h in order to induce dental pulp inflammation. After 24 h, pulp capping was performed and glass
ionomer restorations were placed after capping procedures. (b) Clinical illustration of the pulp exposure site
after 24 h (arrows). After 24 h of pulp exposure, there was no bleeding noted before the pulp capping
procedures. (c) A clinical photo showing glass ionomer restoration at the exposure sites (arrows)
Fig. 6 The center of injury on micro-CT. The vertical and horizontal diameters were determined in the micro-
CT scan slide that shows the largest pulp exposure site. This slide represents the center of the injury. Vertical
diameter represented the diameter of the bur. Horizontal diameter represented the depth of the bur
Fig. 7 Comparison of the results of micro-CT (a) and histology (b) slides. Outline of reparative dentin on micro-
CT slides was referenced to the histology results (red outline); x denotes the pulp exposure site. Bars (b):
300 μm
4. For EDTA preparation: weigh 186.12 g of EDTA·Na2·2H2O

(MW: 372.24). Add the powder into 800 mL water. Add 10 N
NaOH to adjust pH to 8. Make up to 1 L solution with water.
Sterilize and stock at 4 C.
5. The recommended dosage for rats is 40–90 mg/kg of keta-
mine and 5–10 mg/kg of xylazine (administered intraperito-
neally (IP)).
6. Micro-CT scan settings were voxel size 18 μm, 70 kVp, 114 μA,
0.5 mm AL filter and integration time 500 ms. Perform
subsequent analysis using the manufacturer’s evaluation
software.
7. In case you want to compare between two pulp capping mate-
rials, we recommend using a split-mouth design, with each site
treated with a different material in each rat.
8. Compare the micro-CT scan slices to the histology slides and
outline the region of pulp exposure manually slice by slice using
the MiViewer software.
9. Additional histomorphometry: we also recommend staining
the samples with Alcian Blue stain to evaluate reparative dentin
formation and mineralization.
10. Immunohistochemistry (IHC): we recommend conducting an

IHC staining to detect the expression of dentin sialoprotein
(DSP) and interleukin-6 (IL-6) within the dental pulp. The
results could be utilized to determine the activity of osteoblasts
dental pulp differentiation markers (DSP) and the pulpal
inflammation status (IL-6) after pulp injury. Carry out the
immunohistochemical staining according to the manufac-
turer’s instruction of the antibody system of choice using
primary antibodies against DSP and IL-6. We recommend
using 3,30 diaminobenzidine (DAB) as the chromogen.
References
1. Touger-Decker R, van Loveren C (2003) analysis and trial sequential analysis. Clin Oral
Sugars and dental caries. Am J Clin Nutr Investig 20:1121–1132
78:881S 11. MÖLLER ÅJR, FABRICIUS L, DAHLÉN G
2. Koo H, Falsetta ML, Klein MI (2013) The et al (1981) Influence on periapical tissues of
exopolysaccharide matrix: a virulence determi- indigenous oral bacteria and necrotic pulp tis-
nant of cariogenic biofilm. J Dent Res sue in monkeys. Eur J Oral Sci 89:475–484.
92:1065–1073. https://doi.org/10.1177/ https://doi.org/10.1111/j.1600-0722.1981.
0022034513504218 tb01711.x
3. Marsh PD (2006) Dental plaque as a biofilm 12. Schwendicke F, Stolpe M (2014) Direct pulp
and a microbial community—implications for capping after a carious exposure versus root
health and disease. BMC Oral Health 6(Suppl canal treatment: a cost-effectiveness analysis. J
1):S14 Endod 40:1764–1770. https://doi.org/10.
4. Marsh PD, Head DA, Devine DA (2015) Den- 1016/j.joen.2014.07.028
tal plaque as a biofilm and a microbial commu- 13. Sakamoto M, Rôças IN, Siqueira JF, Benno Y
nity—implications for treatment. J Oral Biosci (2006) Molecular analysis of bacteria in asymp-
57:185–191 tomatic and symptomatic endodontic infec-
5. Klein MI, Hwang G, Santos PHS et al (2015) tions. Oral Microbiol Immunol 21:112–122.
Streptococcus mutans-derived extracellular https://doi.org/10.1111/j.1399-302X.
matrix in cariogenic oral biofilms. Front Cell 2006.00270.x
Infect Microbiol 5:10. https://doi.org/10. 14. Siqueira JF, Rôças IN (2009) Diversity of end-
3389/fcimb.2015.00010 odontic microbiota revisited. J Dent Res
6. Marsh PD (1994) Microbial ecology of dental 88:969–981. https://doi.org/10.1177/
plaque and its significance in health and dis- 0022034509346549
ease. Adv Dent Res 8:263–271 15. Fabricious L, Dahlen G, Öhman AE, Möller
7. González-Cabezas C (2010) The chemistry of AJR (1982) Predominant indigenous oral bac-
caries: remineralization and demineralization teria isolated from infected root canals after
events with direct clinical relevance. Dent Clin varied times of closure. Eur J Oral Sci
N Am 54:469–478 90:134–144. https://doi.org/10.1111/j.
8. Neel EAA, Aljabo A, Strange A et al (2016) 1600-0722.1982.tb01536.x
Demineralization–remineralization dynamics 16. Loesche WJ, Gusberti F, Mettraux G et al
in teeth and bone. Int J Nanomedicine (1983) Relationship between oxygen tension
11:4743–4763 and subgingival bacterial flora in untreated
9. Siqueira JF, Rôças IN (2005) Exploiting human periodontal pockets. Infect Immun
molecular methods to explore endodontic 42:659–667
infections: Part 1: Current molecular technol- 17. Ando N, Hoshino E (1990) Predominant obli-
ogies for microbiological diagnosis. J Endod gate anaerobes invading the deep layers of root
31:411–423 canal dentine. Int Endod J 23:20–27. https://
10. Schwendicke F, Brouwer F, Schwendicke A, doi.org/10.1111/j.1365-2591.1990.
Paris S (2016) Different materials for direct tb00798.x
pulp capping: systematic review and meta- 18. Raetz CRH, Whitfield C (2002) Lipopolysac-
charide endotoxins. Annu Rev Biochem
71:635–700. https://doi.org/10.1146/ 20. Levin LG, Rudd A, Bletsa A, Reisner H (1999)

annurev.biochem.71.110601.135414 Expression of IL-8 by cells of the odontoblast
19. Goldberg M, Farges JC, Lacerda-Pinheiro S layer in vitro. Eur J Oral Sci 107:131–137.
et al (2008) Inflammatory and immunological https://doi.org/10.1046/j.0909-8836.1999.
aspects of dental pulp repair. Pharmacol Res eos107209.x
58:137–147. https://doi.org/10.1016/j. 21. Schmalz G, Galler KM (2011) Tissue injury
phrs.2008.05.013 and pulp regeneration. J Dent Res 90:828–829
Part VI
Protocols for Genetic, Epigenetic and Clinical Studies

Chapter 36
Protocol GenoDENT: Implementation of a New NGS Panel

for Molecular Diagnosis of Genetic Disorders with Orodental
Involvement
Tristan Rey, Julien Tarabeux, Bénédicte Gerard, Marion Delbarre,
Antony Le Béchec, Corinne Stoetzel, Megana Prasad,
Virginie Laugel-Haushalter, Marzena Kawczynski, Jean Muller,
Jamel Chelly, Hélène Dollfus, Marie-Cécile Manière,
and Agnès Bloch-Zupan
Abstract
Rare genetic disorders are often challenging to diagnose. Anomalies of tooth number, shape, size, miner-
alized tissue structure, eruption, and resorption may exist as isolated symptoms or diseases but are often
part of the clinical synopsis of numerous syndromes (Bloch-Zupan A, Sedano H, Scully C. Dento/oro/
craniofacial anomalies and genetics, 1st edn. Elsevier, Boston, MA, 2012). Concerning amelogenesis
imperfecta (AI), for example, mutations in a number of genes have been reported to cause isolated AI,
including AMELX, ENAM, KLK4, MMP20, FAM83H, WDR72, C4orf26, SLC24A4, and LAMB3. In
addition, many other genes such as DLX3, CNNM4, ROGDI, FAM20A, STIM1, ORAI1, and LTBP3 have
been shown to be involved in developmental syndromes with enamel defects. The clinical presentation of
the enamel phenotype (hypoplastic, hypomineralized, hypomature, or a combination of severities) alone
does not allow a reliable prediction of possible causative genetic mutations. Understanding the potential
genetic cause(s) of rare diseases is critical for overall health management of affected patient. One effective
strategy to reach a genetic diagnosis is to sequence a selected gene panel chosen for a determined range of
phenotypes. Here we describe a laboratory protocol to set up a specific gene panel for orodental diseases.
Key words Genetic variations, High-throughput sequencing, Gene panel, Probe enrichment, Liquid
capture, Mendelian disorders, NextSeq 550, Dental anomalies, Genetics, Syndromes, Rare diseases
Abbreviations
BR Broad range
CNV Copy number variation
DMSO Dimethyl sulfoxide
DNA Deoxyribose nucleic acid
dNTP Deoxynucleoside triphosphate
dsDNA Double-stranded DNA
407
408 Tristan Rey et al.
EDTA Ethylenediaminetetraacetic acid

gDNA Genomic DNA
HS High sensitivity
HTS High-throughput sequencing
IQC Internal quality control
MAF Minor allele frequency
NGS Next-generation sequencing
PCR Polymerase chain reaction
RNase Ribonuclease
RT Room temperature
SNP Single-nucleotide polymorphism
SNV Single-nucleotide variant
1 Introduction
Many diseases have clear etiological links to genetic disorders.

Understanding which pathogenic gene variants cause these defects
may change patient’s care. Rare genetic disorders are difficult to
diagnose, their rarity making research and knowledge diffusion
arduous. To address this problem, the RARENET project
(http://www.rarenet.eu/en/) congregates partners from different
European regions. Indeed, RARENET is a French-German-Swiss
cross-border cooperative project, in the framework of
INTERREG V, EU (FEDER)-funded program, which has the
objective of improving the management and health of patients
with complex rare diseases. RARENET brings together many pro-
fessionals investigating specific types of rare disease (rare diseases
presenting with orodental anomalies and/or autoimmune defects),
providing a platform of interaction, education, and resources. A
major advantage of a large grouping is improved access to patient
data, thus enhancing patient cohort size and rare disease data
collection. Overall, improving and increasing the number of clinical
descriptions of rare disease gene variants will positively impact
patient’s health. The GenoDENT NGS (next-generation sequenc-
ing) tool aims to improve knowledge and molecular diagnosis of
rare genetic diseases with orodental manifestations. One of the
current strategies for rare disease analysis is to use a NGS panel.
This applies a screen for numerous selected genes to decipher which
is a responsible variant. Here we will detail the setup of the NGS
GenoDENT project in the RARENET initiative.
Application of a new diagnostic technique, from its use in a
research setting to its practical application in clinical diagnosis, is a
long process and possesses a variety of challenges. Utmost care is
devoted to ethical management of patient privacy, so the private
health care information is respected and protected. Patients should
be well aware of why and how their biological samples are used;
Protocol GenoDENT 409
hence clear explanations and patient informed consent are required

before any study. The stepwise procedure to establish a NGS panel
starts from research data in the field to design the panel, followed
by specific applications of the panel to a more routine clinical
diagnosis. Steps include gene selection, panel design for NGS,
patient sample collection, sample coding for anonymity, DNA
extraction and sequencing, a bioinformatics pipeline, results ana-
lyses, and diagnosis delivery.
2 Materials
2.1 Structures – Public hospital-university complex (agreement, partnership).

– Experimented clinicians and researchers in the field.
– Rare disease centers.
– Operational accredited laboratory for DNA extraction and storage,
NGS experiments and analysis (see Note 1), and official report.
2.1.1 CRMR O-Rares France is a leading country in rare diseases management, having
(Reference Center for Rare launched two large sequential programs for rare diseases, since
Oral Diseases) 2004. The first program has two funded national reference centers
focusing on orodental manifestations of rare diseases. The Stras-
bourg reference center was recognized in 2006, reaccredited in
2011 after HAS (French High Health Authority) assessment, and
installed in 2017 as the coordinating French Reference Centre
(CRMR) for rare oral and dental diseases (O-Rares), located at
the Hôpitaux Universitaires de Strasbourg. O-Rare is one of the
leading reference centers, cooperating with other groups (Paris
Rothschild, constitutive site, along with 16 affiliated competence
centers throughout France). The network welcomes several groups
of patients. It investigates and manages patients from childhood to
adulthood stages, presenting rare developmental, either isolated or
syndromic, orodental anomalies. There is a specific clinical track for
hypodontia/oligodontia patients, dentinogenesis patient, and/or
amelogenesis imperfecta patient, as these patients require long-
term complex multidisciplinary treatments. Among these, patients
and families presenting unique phenotypes are selected for further
genetic and clinical investigation. Understanding rare disease
patients presents specific orodental multidisciplinary management
needs (prevention, sedation, general anesthesia), related to the full
scope of their clinical problems (including neurologic, sensorial,
age-related problems). Clinicians also address rare disease patients
presenting nongenetic oral manifestations, such as caries, periodon-
tal diseases, and trauma.
Our expertise mission, defined by Rare Disease National Plan,
is to act as a tertiary center from diagnosis to treatment (referral
center) for other health professionals. Our center takes care of

800 patients per year. The center collaborates with other refer-
ence/competence centers and hospital units (genetics, pediatrics,
dermatology, maxillofacial, etc.). Our center sets up a child to adult
transition for the follow-up of former pediatric patients.
The center developed the genetic diagnosis (GenoDENT) for
all orodental anomalies.
For training activities, the center provides regional, national,
and international expert training (university, continuing medical/
dental information, patients) and has set up national and interna-
tional guidelines. Furthermore, the reference center is very active in
its liaison with affiliated patient support groups, in initial, postgrad-
uate, and continuous education for various health professionals
(dentists, physicians, geneticists, etc.).
Employing an Internet-based approach—the www.phenodent.
org D4/phenodent biomedical database—we have been able to
unite clinicians in a securely controlled accessible database which
has shown to be an effective tool to identify new patient cohorts
with oral manifestations of rare diseases. From a cohort of over
4000 patients, we have screened affected individuals and families
presenting a range of oral and teeth developmental defects. Our
clinical team begins patient screenings with characterization of
pathologies and directed treatment. Typical dental malformations
include abnormalities of tooth number (hypodontia, oligodontia,
anodontia), shape, size, and defects in hard tissues, i.e., enamel
(amelogenesis imperfecta), dentin (dentinogenesis imperfecta,
dentin dysplasia), cement, alveolar bone formation and mineraliza-
tion, disrupted tooth eruption, tooth resorption, and cleft lip/pa-
late. Orodental findings are documented using the D[4]/
phenodent Diagnosing Dental Defects Database registry (www.
phenodent.org) which is approved by CNIL (French National
commission for informatics and liberty, number 908416).
We implemented a biological sample collection (saliva, DNA,
tissues, teeth, etc.) “Orodental Manifestations of Rare Diseases”
registered within the MESR (French Ministry of Higher Education
and Research) Bioethics Commission DC-2012-1677 within
DC-2012-1002CPP (person protection committee) Est IV on
11/12/2012.
The reference center leads national (PHRC) and international
(INTERREG) clinical research programs (e.g., data collection/
phenotyping (D4/phenodent) and biosample collection (mainly
saliva samples) for genotyping and photographic database
(MediaD4). Regulatory approvals have been obtained, and proce-
dures follow clinical practice guidelines (information, consent,
recording in medical file, anonymity, right to claim data). All
other clinical research (therapeutic, observational) projects con-
ducted temporarily in the reference center are subject to specific
information and consents. Each treating practitioner of the
reference center carefully explains the potential benefits/risks of all

procedures (treatment, test) to all patients.
Clinical research activities are registered in ClinicalTrials.gov:
NCT01746121 and NCT02397824. A. Bloch-Zupan is within the
CRMR O-Rares in charge of research activities and developing the
D4/phenodent database, the GenoDENT genetic diagnostic tool.
2.1.2 Head and Neck The second rare disease French program has built up a potent
National Network network (so-called Filière Nationale TETECOU). This network
includes 5 reference centers, 35 competence/17 expert centers,
and 21 French patient support groups. Groups concentrate in the
field of rare diseases (also involved in EURORDIS). The network
acts in collaboration with national French scientific societies. Taken
together, this network includes over 50,000 patients with rare
craniofacial, oral, and ear, nose, and throat (ENT) disorders.
2.1.3 RARENET Is a tri-national network for education, research, and management

of complex and rare diseases in the Upper Rhine region. The
RARENET network (INTERREG V) aims to improve the manage-
ment of patients with autoimmune and orodental rare diseases by
the interconnection, interaction, education, and information of
different target groups such as practitioners, researchers, teachers,
students, operational staff (technicians, research engineers, clinical
research associates), health professionals (private and public sec-
tors), patients and patient support groups, and the general public.
Bringing together reference centers for rare diseases, universities,
research centers, and health industries, RARENET offers this plat-
form for education and interaction directly linking clinical treat-
ment to the needs of patients, which are represented by their
respective support groups. Our innovative actions include interdis-
ciplinary training, applied training for health professionals, infor-
mation initiatives for both patients and the general public,
international networking of all participants, sharing a collection of
biological samples and genetic data, and developing a distance-
learning tool. Project duration is 4 years (01/01/2016–31/12/
2019). Total budget is 3,979,174 €. EU funding is 1,989,587 €
(European Regional Development Fund (ERDF)-INTERREG V
Upper Rhine program). A Bloch-Zupan is the project leader for the
RARENET program.
2.1.4 ERN CRANIO 2017 The best medical centers and universities across Europe are con-
necting into 24 networks composed of nearly a thousand centers of
expertise or highly specialized health-care providers. The CRMR
O-Rares is a partner of the ERN CRANIO and in charge of the
orodental disease track.
2.2 Saliva Collection – 15 mL compatible centrifuge (up to 3500 g).

– Kit prepITlL2P (DNA Genotek ref : PT-L2P-5).
– Water bath (up to 50 C).
– 15 mL Falcon tube.
– Ethanol (100%).
– DNA storage buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.0).
– Optional: 5 mL sterile plastic syringe.
– Optional: sterile forceps.
2.3 Qubit – Kit: dsDNA BR Assay (ref: Q32850) (Thermo Fisher

Quantification Scientific®).
– Kit dsDNA HS Assay (ref: Q32851) (Thermo Fisher Scientific®).
– Qubit 2.0 (Thermo Fisher Scientific®).
2.4 NGS Library l LabChip GX Caliper (PerkinElmer®).

Preparation l Genomic DNA Reagent Kit (PerkinElmer®).
l HT DNA Extended Range Chip (PerkinElmer®).
l 2100 Bioanalyzer Instruments (Agilent Genomics®).
l Agilent DNA 1000 Kit (Agilent Genomics®).
l Agilent High Sensitivity DNA Kit (Agilent Genomics®).
l Vortex mixer IKA—Model MS3.
l Thermocycler.
l SureSelect QXT Reagent Kit for NextSeq (ref: G9683B) (Agi-
lent Genomics®).
l SureSelectXT Custom 0.5–2.9 Mb, 96 (ref 5190-4817) (Agi-
lent Genomics®).
l Dynabeads™ MyOne™ Streptavidin T1 (ref: 65602) (Thermo
Fisher Scientific®).
l AMPure XP (Agencourt®).
l Magnetic rack.
l LoBind tubes.
2.5 High-Throughput l NextSeq 550 (Illumina®).

Sequencing l NextSeq 500/550 Mid Output v2 kit (300 cycles)
FC-404-2003 (Illumina®).
l 10 N NaOH.
l 1 M Tris–HCl, pH 7.0.
2.6 Identity l Applied Biosystems™ Pre-designed SNP genotyping assays,

Monitoring human, Small-Scale Applied Biosystems (ref41106302).
l TaqMan™ Genotyping Master Mix 2X (ref: 4371355).
l LightCycler® 480 Instrument plate 96, white (ref: 04 729
692 001).
l LightCycler® 480 Sealing Foil (ref: 04 729 757 001).
l LightCycler® 480 Instrument, 96-well (ref: 05 015 278 001).
3 Methods
3.1 Gene Selection All start with a review of field-related knowledge by a competent
clinical scientist. The important genes or associated regions for a
phenotype of interest are selected [2]. Out of this selection process,
a gene list is created with two distinct groups. The first one contains
genes with referenced literature reporting proof of pathogenicity in
human diseases. The second one contains suspicious genes and
genes in the literature, which are reported pathogenic in animal
models (see Notes 2 and 3).
3.2 Panel Design Probes for NGS sequencing can be designed using the Agilent
SureDesign portal (https://erray.chem.agilent.com/suredesign,
Agilent, USA). When registered, one can log in and go to the
SureSelect DNA section. Here you can list your selected gene,
and the software will design probes according to selected
parameters.
l Log in with email and password.
l Select “SureSelect DNA,” and tick the box “Show Advanced
Options” (Fig. 1).
l Tick “Advanced” and “Create design” (Fig. 2) and click
“continue.”
Fig. 1 Part of the SureDesign home screen (https://erray.chem.agilent.com/suredesign, Agilent, USA)

Fig. 2 Selected parameters to start a SureSelect DNA design on the Agilent SureDesign portal (https://erray.
chem.agilent.com/suredesign, Agilent, USA)
Fig. 3 Selected parameters to search region of interest for design on the Agilent SureDesign portal (https://
erray.chem.agilent.com/suredesign, Agilent, USA)
l Name the design. Adding a description and keyword is optional.

Click “next.”
l Tick “Design new probes by tiling genes or regions,” and click
“next.”
l Enter the design targets; select the database and details about
the region of interest (Fig. 3).
l Click “next.”
l The SureDesign tool summarizes which target is found or not
and details about the target (number of regions, length, and
position).
l Click “next.”
l Select probe design parameters (density, masking, boosting,
extension into masked, strand).
We recommend to use (see Note 4, Fig. 4):
Fig. 4 Selected parameters for the probe design of the region of interest on the
Agilent SureDesign portal (https://erray.chem.agilent.com/suredesign, Agilent,
USA)
l At least a tiling density of 2.

l The least stringent masking parameters.
l The maximize performance option.
l An extension into repeats of 20.
l Using the sense strand.
l Extending coding exon region by 25 bases from 30 end and
25 bases to 50 .
l Click “Begin Probe Selection.”
After you send your request, the software will generate design
data which can be directly downloaded (see Note 5). Once the
design is complete, you can order it online.
The GenoDENT 4.0 panel contains 513 genes (see Annex 1)
which represent a capture size of 2.365 Mbp targeted by 57,670
probes.
3.3 Centralization of A communication and sensitization about the project and what the
Relevant Patient new diagnostic solution has to offer must be disseminated to all the
Sample (Spread medical partners who will recruit relevant patients. Patient inclu-
Information (Who, sion criteria must be well defined. Clinical staff of the participating
How, Sample institutions will recruit patients who match the inclusion criteria for
Condition Expedition) voluntary participation onto the research project. Moreover, clin-
Anonymity, Patient icians must have an easy access to instruction for the bio-specimen
Consent, Data Secure sampling (saliva in OG-250 Oragene DNA®, DNA Genotek,
Storage) Canada, min 3 mL; blood in EDTA, etc.) from the participants
and the selected family relatives (affected, non-affected). For rele-
vant genetic diagnosis, patients’ (affected individuals) and parents’
samples are required. The required documentation includes sample
delivery procedures and patient/relative consent forms. The con-
sent forms are mandatory for any analysis and have to be correctly
filled out, with signatures of patients and the treating clinician
collected in the presence of a witness. If the patient is a minor,
one or both parents can sign. The child, if able, will also give
consent. The project can provide/send sampling kits like
RARENET with OG-250 (Oragene DNA®, DNA Genotek Inc.,

Ottawa, Canada). The contact information of the researcher clini-
cian in charge is available, in case problems arise.
In order to simplify and centralize clinical research data, a web
support is required. This web platform allows standardized collec-
tion of clinical disease data, linked to patient medical history and
global multidisciplinary family care information. Included in this
setup is epidemiologic monitoring, to evaluate the quality, efficiency
(morbidity), and medico-economic consequences of the medical
care. Moreover, a larger clinical data set is a powerful tool. This
web platform will help toward knowledge organization and diffusion
and promote improved quality research. Clinicians, through this
platform, have access to a medical support of specialists in the field.
The web support for the GenoDENT project is within the D[4]/
phenodent (http://www.phenodent.org/indexgb.php) platform. In
this platform, with required consent, patients’ anonymized data and
samples can be used for research and genetic testing activities.
3.4 DNA Extraction, In our application, samples consist of mostly saliva. This sampling
Reference, and Storing method has some advantages: noninvasive, relatively easy to per-
form, and samples’ stability for years at room temperature (RT).
3.4.1 DNA Extraction l Check at arrival if the sample is correctly closed.

from Saliva l Weigh the container to determine the saliva quantity (g) (the
empty OG-250, OG-500, and OG-575 (Oragene DNA®, DNA
Genotek, Canada), respectively, weight 14.15 g, 6.81 g, and
5.66 g (Fig. 5)).
l There is no need to weight the OCR-100 kit.
l Mix the sample by repeated rolling and shaking up the sample
for few seconds.
Fig. 5 Different available kits to collect saliva: here illustrated Oragene. DNA OG-250 DNA-Genotek Inc.
Ottawa, Canada, weight ¼ 14.15 g; OG-500, weight ¼ 6.81 g; OG-575, weight ¼ 5.66 g
l Incubate the sample in a water bath at 50 C for 1 h to overnight

(see Note 6).
l [Optional procedures if sponges are present, after 50 C incubation

step
l Remove all the saliva that you can from the kit, and transfer it to a
15 mL Falcon tube.
l Place the barrel of a 5 mL plastic syringe (without the plunger)
into the Falcon tube.
l Transfer the sponges with sterile forceps into the barrel (Fig. 6),
and centrifuge the Falcon tube at 200 g for 10 min at RT.
l The saliva contained in the sponge will reach the Falcon tube. Then
you can discard the barrel and the dry sponges and continue the
protocol as with other samples.]
l Once the saliva is fluid, transfer it into a 15 mL Falcon tube with

the patient name and DNA number and note “saliva” on
the tube.
l Record the transferred volume and add 1/25 of the total volume
of prepITlL2P.
l Vortex the Falcon tube and store it at 20 C for 10 min.
l Centrifuge the solution at RT and 3500 g for 20 min.
l Transfer the supernatant to a new 15 mL Falcon tube; include an
identical identification.
l Leave a little volume of supernatant with the pellet to ensure that
any of the turbid impurities are taken out. Discard the Falcon
tube with the pellet.
Fig. 6 Insertion of a filled up 5 mL plastic syringe barrel with sponges into a

15 mL Falcon tube
l Add a volume of ethanol (95–100%) equivalent to 1.2 times of

the supernatant volume to the Falcon tube.
l Gently mix by 10 inversions and leave the sample at RT for
10 min. At this step, a white clot of DNA fiber must be visible.
If the clot is not visible, an incubation at 20 C for 20 min may
be considered to help the DNA to precipitate.
l Repeat the centrifuge step (3500 g, 10 min, RT). At this point
DNA should be visible forming a pellet and a trail along
the tube.
l Carefully remove the supernatant with a pipette set on the other
side of the trail.
l Add slowly 1 mL of ethanol 70% at RT without disturbing the
pellet and the trail.
l Incubate the sample 1 min at RT.
l Agitate gently the tube, and remove completely the ethanol
without disturbing the pellet and the trail. If the pellet pulls
away, repeat the centrifugation step for 5 min.
l Rehydrate the DNA with 200 μL to 1 mL DNA storage buffer
(10 mM Tris–HCl, 1 mM EDTA, pH 8.0, e.g., elution buffer
AE Qiagen®) depending on the importance of the DNA fiber
clot, and vortex for 30 s.
l Incubate the extracted DNA in a water bath at 50 C for 1 h to
overnight if necessary, and gently vortex regularly.
l At this point, transfer the DNA to a microtube labeled with
patient’s first name, last name, and DNA number; it is ready
for the quantification step.
l DNA can be stored at 4 C or 20 C. If DNA is stored at
20 C for a long time, avoid freeze-thaw cycles.
3.5 DNA The quantification is performed using a fluorescent method that

Qualification allows quantifying only dsDNA (dsDNA BR Assay). The quantifi-
cation requires calibration points (standards 0 and 100 μg/L)
3.5.1 Quantification
included in the kit.
Using Qubit BR
l Prepare a mix for n + 4 tubes (n ¼ number of sample) using
199 μL of dsDNA BR buffer for 1 μL of dsDNA BR reagent (see
Notes 7 and 8) (Table 1).
l Dilute your standard and sample into this fluorescent dilution
(Table 2), and vortex (see Note 9).
l Select DNA dsDNA broad range (BR).
l Measure first the standard 1 and then the standard 2. Then, a
DNA with known concentration is used to confirm that calibra-
tion is appropriate.
l Finally measure the samples.
Table 1
Mix for the fluorescent dilution for Qubit BR quantification
Reagents For 1 tube

dsDNA BR buffer 199 μL
dsDNA BR reagent 1 μL
Total 200 μL
Table 2
Final mix for Qubit BR quantification (standard, samples, and control)
Mix for Final mix

Standards 10 μL standard + 190 μL fluorescent dilution
Sample 10–50 ng/μL 10 μL sample + 190 μL fluorescent dilution
DNA control Same dilution as your samples
3.5.2 DNA Quality The global quality of the DNA (absence of DNA degradation) is
verified with LabChip GX (PerkinElmer®) using a genomic chip.
Before starting your library preparation, it is important to
investigate the DNA profile of the samples. A genomic chip with
the Genomic DNA Reagent Kit (PerkinElmer®) will migrate your
samples through a gel by electrophoresis. For the following proto-
col, your reagent must be at RT, so remove them from the refriger-
ator (4 C) prior to Genomic DNA Reagent Kit and chip testing.
Gel-Dye Preparation l Add 13.75 μL of DNA dye concentrate (blue lid) to a tube of
Genomic DNA Gel Matrix (red lid).
l Vortex vigorously.
l Transfer all the gel into two spin filters (approximately
550 μL each).
l Centrifuge for 10 min at 9200 g.
l Discard filter and note the date on the tube.
Ladder and Zipper Wash The gel can be stored at 4 C in the dark for up to 3 weeks.
Tube Preparation
l In a 0.2 mL tube (provided in the kit), add 12 μL of DNA ladder
(yellow lid), without vortexing, in 108 μL of fresh ultrapure
water.
l Mix by pipetting up and down.
Fig. 7 LabChip GX Caliper (PerkinElmer®) interface
l Add 750 μL of ultrapure water in a 0.75 mL tube (provided in

the kit).
l On the caliper, press “Eject” and place both tubes in their place.
l Press “Eject” again to close the front door as in (Fig. 7).
Sample Preparation l Add 100 ng of Genomic DNA diluted in 30 μL of ultrapure

water in the microplate.
l Seal the plate with an adherent film, and centrifuge the plate for
10 min at 1000 g.
l Carefully remove the film, and place the microplate on the
microplate carrier as in Fig. 8.
Chip Setup For 1–24 samples (Fig. 9):

l Remove reagent from the chip using a vacuum.
l Wash by adding 200 μL of ultrapure water in the wells 1, 3, 4, 7,
8, and 10.
l Vacuum the wells, and repeat this washing step minimizing the
time when wells are dry. Check if there is no remaining water
using the provided cotton bud.
l Add 60 μL of Genomic DNA Marker (green lid) in the well 4.
l Using a reverse pipetting technique, add 50 μL of gel-dye to the
wells 3, 7, 8, and 10.
For 25–48 samples (Fig. 9):
l Remove reagent from the chip using a vacuum.
Fig. 8 Labchip GX Caliper open door with microplate, ladder, and buffer vial on
Fig. 9 Schema of genomic chip for LabChip GX Caliper with annotated well
number
l Wash by adding 200 μL of ultrapure water in the wells 1, 3, 4, 7,

8 and 10.
l Vacuum the wells and repeat this washing step minimizing the
time when wells are dry. Check there is no remaining water using
the provided cotton bud.
l Add 120 μL of Genomic DNA Marker (green lid) in the well 4.
l Using a reverse pipetting technique, add 75 μL of gel-dye to the
wells 3, 7, and 8. For the well 10, add 120 μL of gel-dye.
Fig. 10 Closed chip compartment of the LabChip GX Caliper
Fig. 11 Opened chip compartment of the LabChip GX Caliper
l Check if the chip window is clean; if not use a precision wipe

with 70% isopropanol to clean it (see Note 10).
l To load the chip, press “Chip,” open the lid latch (Fig. 10), and
place the chip in his compartment (Fig. 11).
l Close the lid and be sure to clip the lid latch.
l Delicately push the chip track into its initial position.
l Indicate on the appearing windows in the software what kind of
chip you are using: gDNA.
Starting Run l On the LabChip GX software, click “Instrument” and “Start
Prime.” The priming lasts for 10 min (see Note 11).
Table 3
Mix for the fluorescent dilution for Qubit HS quantification
Reagents For 1 tube

dsDNA HS buffer 199 μL
dsDNA HS reagent 1 μL
Total 200 μL
l On the Caliper click “Eject” to load your sample plate.

l Click “Eject” again to close the front door.
l On the software click Run and enter your parameters: select the
zipper at 3 mm, and name your run in the “output” tab.
l Click “Start” (see Note 12).
3.5.3 Quantification The quantification is performed using a fluorescent method that

Qubit HS allows quantifying only dsDNA (dsDNA HS Assay). The quantifi-
cation requires calibration points (standards 0 and 10 μg/L)
included in the kit.
l Prepare a mix for n + 4 tubes (n ¼ number of sample) using
199 μL of dsDNA BR buffer for 1 μL of dsDNA BR reagent (see
Notes 7 and 8) (Table 3).
l Dilute your standard and sample into this fluorescent dilution,
and vortex (see Note 9).
l Select DNA, dsDNA high sensitivity (HS).
l Measure first the standard 1 and then the standard 2. Then a
DNA with known concentration is used to confirm that calibra-
tion is appropriate.
l Finally measure the samples.
3.6 NGS Library l Dilute DNA (from data of Qubit HS quantification) at

Preparation Using [18–25] ng/μL in a LoBind tube 1.5 mL with ultrapure water.
Agilent SureSelect QXT l Quantify again the final dilution with Qubit HS.
Protocol l Readjust the dilution if it does not reach the range (see Note 13).
3.6.1 Library
Hybridization
DNA Dilution
Fragmentation and Ligation During this step, the gDNA will be fragmented with enzyme, and
of Adapters: Pre-PCR Zone adapters will be ligated to the ends in one reaction.
Mixes are made under a hood in LoBind tube of 1.5 mL.
Keep the mix on ice or in a cold rack.

For the first use of the kit, add 1.5 mL 100% ethanol in the
bottle with 4.5 mL of Stop Solution. Final concentration of ethanol
will be 25%. Sign/date the bottle to note ethanol addition.
Take out reagent from the 20 C (SureSelect QXT Enzyme
MIX ILM, SureSelect QXT Buffer, and DMSO), and leave them on
ice or at 4 C (except the DMSO). At least 30 min before use, take
out the AMPure beads.
Prepare 75% ethanol for the purification steps.
Check if the SureSelect QXT Stop Solution has ethanol inside.
Turn on the thermocycler (with heating lid).
l Vortex SureSelect QXT Enzyme MIX ILM and SureSelect QXT
Buffer.
l Prepare the mix on a cold rack; add reagent in the order.
l In each well add 17 μL of SureSelect QXT Buffer.
l Then add 2 μL of the diluted DNA (max 25 ng/μL) in tubes.
l Add 2 μL of SureSelect QXT Enzyme MIX ILM in each, and mix
up and down 8–10 times.
l Close tubes, vortex 20s, and quickly centrifuge.
l Place immediately your samples in the thermocycler and run
(Table 4) for 10 min at 45 C and 1 min at 4 C and hold at 4 C.
l After the 1 min incubation, immediately place samples on a cold
rack, and add 32 μL of Stop Solution in each. Change the
barrette lead.
l Vortex and centrifuge briefly, leave the samples for 1 min at RT,
and proceed to the purification steps.
Purification with AMPure Take out the beads 30 min before the experiment and leave them at
Beads RT. Vortex roughly to resuspend the beads until the reagent is
homogeneous.
l Add 52 μL of beads to the reaction mix. Mix up and down.
l Incubate for 5 min at RT and centrifuge.
Table 4
Thermocycler program for fragmentation and ligation of adapters
Temperature ( C) Time
45 10 min
4 1 min
4 1
l Place sample on the magnetic rack and wait for 2 min. The
supernatant has to be clear.
l Leaving samples on the magnetic rack, take out the supernatant
without touching the beads (see Note 14).
l Add 90–100 μL of 75% ethanol in each sample.
l Incubate for 1 min at RT and then remove the ethanol.
l Repeat this washing step with ethanol one time.
l Spin off the sample and replace them on the magnetic rack.
l Take out the remaining ethanol with a 1–10 μL pipette.
l Let the sample dry on the magnetic rack about 2 min at RT, lid
open with Parafilm on tube (see Note 15).
l Suspend the beads in 11 μL of ultrapure water. Mix up
and down.
l Incubate for 2 min at RT and then centrifuge.
l Replace sample on the magnetic rack, and wait for the superna-
tant to become clear (around 2 min).
l Take out the supernatant (around 10 μL), and transfer it in a
new PCR tube on a cold rack (4 C) (see Note 16).
Library Amplification l Prepare the library amplification mix (Table 5), mixing for each
reaction: 25 μL of ultrapure water, 10 μL of 5 Herculase II
Reaction Buffer, 0.5 μL of 100 mM dNTP Mix, 2.5 μL of
DMSO, 1 μL of SureSelect QXT Primer Mix, and 1 μL of
Herculase II fusion DNA Polymerase.
l Vortex thoroughly the mix and centrifuge.
l Distribute 40 μL of this mix in each tube containing 10 μL of
purified sample.
l Then vortex and centrifuge.
Table 5
Library amplification mix
Reagents Volume for 1 reaction (μL)

Ultrapure water 25
5 Herculase II Reaction Buffer 10
100 nM dNTP Mix 0.5
DMSO 2.5
SureSelect QXT Primer Mix 1
Herculase II fusion DNA Polymerase 1
Total volume 40
Table 6
Library amplification program
Cycle number Temperature ( C) Time

1 68 2 min
1 98 2 min
8 98 30 s
57 30 s
72 1 min
1 72 5 min
1 4 Hold
l In a thermocycler incubate (Table 6) samples for 2 min at 68 C

and 2 min at 98 C, following by eight loops of 30 s at 98 C,
30 s at 57 C, and 1 min at 72 C. To finish the run, incubate the
samples for 5 min at 72 C, and hold at 4 C.
Purification with AMPure l Take out the beads 30 min before the experiment and leave them
Beads (Intermediate Zone) at RT. Vortex roughly to mix the beads until the reagent is
homogeneous.
l Add 50 μL of beads to the reaction mix. Mix up and down.
l Place sample on the magnetic rack and wait 2 min. The superna-
tant has to be clear.
without touching the beads (see Note 14).
l Incubate for 1 min and then remove the ethanol.
l Centrifuge the sample and replace them on the magnetic rack.
l Let the sample dry on the magnetic rack around 3 min at RT (see
Note 15).
l Mix the beads in 13 μL of nuclease-free water. Mix up and down.
l Incubate for 2 min at RT then centrifuge.
l Take out the supernatant (around 13 μL), and transfer it into a
new PCR tube on a cold rack (4 C) (see Notes 16 and 17).
Fig. 12 Plate base of the Bioanalyzer chip priming station. Set the priming station
on the C position as indicated to use the DNA1000 or the HS chip
Fig. 13 Syringe positions of the Bioanalyzer chip priming station for using the
DNA1000 or the HS chip
Precapture Library A quantification and verification of the precapture library are per-
Quantification formed using an Agilent DNA1000 chip for Bioanalyzer:
l Adjust the plate base of the chip priming station in position C
(Fig. 12).
l Adjust the syringe clip to the lowest position (Fig. 13).
l Incubate the kit at RT for 30 min before using it.
Fig. 14 Picture of DNA1000 chip indicating where to load (well 15) the first 9 μL
of gel-dye before priming the chip
Prepare the gel-dye mix:

l Vortex the DNA dye concentrate (blue lid) for 10 s and
spin down.
l Add 25 μL of DNA dye concentrate into a DNA gel matrix tube
(red lid).
l Vortex the mix and transfer it to a spin filter.
l Centrifuge the spin filter at 2240 g for 15 min at RT.
l Discard the filter and label the tube with the preparation date.
The gel can be used within 4 weeks and stored in the dark at
4 C.
Loading the gel-dye mix:
l Place new DNA1000 chip in the priming station.
l Pipette 9 μL of gel-dye mix at the bottom of the well (Fig. 14).
l Set the timer to 60 s.
l Place the plunger at 1 mL and then close the chip priming
station.
l Be sure that the lock clicks correctly.
l Press the plunger of the syringe down until it is held by the clip.
l Wait 60 s and release the plunger.
l Let the plunger move back until it stops and then slowly pull
back the plunger to the 1 mL position.
l Open the priming station.
l Add 9 μL of gel-dye mix in the other well-marked (Fig. 15).
Fig. 15 Picture of DNA1000 chip indicating where to load the second and third
(well 13 and 14) 9 μL of gel-dye after priming the chip
Fig. 16 Picture of DNA1000 chip for loading of marker, ladder, and samples.
Load 5 μL of marker in well 1 to 12 and 16. Then load 1 μL of ladder in the well
16. Load 1 μL of sample in well 1 to 12 (one sample per well)
Loading the chip:

l Add 5 μL of DNA marker (green lid) into all the wells you did
not add gel mix (ladder included) (Fig. 16).
l Pipette 1 μL of ladder (yellow lid) in the well with a ladder. Add
1 μL of your sample per well or deionized water in the unused
wells (Fig. 16).
l Vortex for 60 s at 2400 rpm in the vortex mixer IKA-Model
MS3.
l Load the chip within 5 min in the Bioanalyzer.
Fig. 17 Selection of the corresponding assay on the Bioanalyzer 2100 Expert Software (Agilent Technologies,
Inc.)
Start the run:

l Select the appropriate assay; click “dsDNA” and then “DNA
1000 Series II.xsy” (Fig. 17).
l Indicate the run data destination.
l Enter you sample name in the sample name table (see Note 19).
Electrode cleaning:
l Wash the electrodes, before and after every run, with the wash
chip loaded with 350 μL of deionized analysis-grade water.
Qualification criteria:
l DNA fragment size should be between 245 and 325 bp.
l A minimum of 750–1500 ng is for a capture design >3.0 Mb
and 500–750 ng for a capture design <3.0 Mb; 500 ng is
required to the hybridization step (see Note 20).
l If the profile is too spread (>1500 bp), dilute the library, and
check them on Bioanalyzer with a HS chip.
Hybridization For a design capture <3.0 Mb:

Vortex all reagents before use.
In strips of PCR tube, add 750 ng up to 950 ng of precapture
library.
l Adjust the volume to 12 μL with ultrapure water.
l Pipette 5 μL pf SureSelect QXT Fast Blocker Mix.
l Mix up and down 10 times, vortex, and spin down.
l Place the barrette in a thermocycler, and incubate (Table 7) for
5 min at 95 C, 10 min at 65 C, and 1 min at 65 C
(STOP step).
l During the STOP step, add the hybridization mix while barrette
is still in the thermocycler.
l Restart the thermocycler, and perform 60 cycles at 60 C for
1 min and 37 C for 3 s. Then hold at 65 C.
Table 7
Thermocycler program for hybridization

1 95 5 min
1 65 10 min
1 65 1 min (STOP)
60 65 1 min
37 3s
1 65 Hold
Table 8
Reagents and volumes per hybridization for the mix of 25% RNAse Block
solution
Reagents Volume for 1 hybridization (μL)

SureSelect RNase Block 0.5
Ultrapure water 1.5
Table 9
Reagents and volumes per hybridization for the hybridization mix
Volume for
Reagents 1 hybridization (μL)
25% RNase Block solution 2
SureSelectXT Custom Capture Library <3 Mb 2
(probes: Oligo baits)
SureSelect QXT fast hybridization buffer 6
Ultrapure water 3
Total volume 13
Hybridization mix:
l On ice prepare a mix of 25% RNAse Block solution with 0.5 μL
of SureSelect RNase Block and 1.5 μL ultrapure water per
hybridization (Table 8).
l Spin down quickly.
l At RT prepare the hybridization mix, keeping probes on ice
(Table 9).
l For each hybridization volume, add 2 μL of 25% RNAse Block

solution, 2 μL of SureSelectXT Custom Capture Library <3 Mb
(probes: oligo baits), 6 μL of SureSelect QXT Fast Hybridiza-
tion Buffer, and 3 μL ultrapure water.
l Vortex and spin down the mix.
l During the STOP step, pause the thermocycler, and without
removing the sample from it, add 13 μL of hybridization mix.
l Mix up and down 10 times and put a new lid.
l Quickly vortex and spin down the strip, and replace it immedi-
ately in the thermocycler.
l Restart the program.
3.6.2 Library Capture First wash the streptavidin beads:

and Amplification
l Vortex the beads and pipette 50 μL in a barrette for each sample.
Library Capture with l Add 200 μL of binding buffer.
Streptavidin Beads l Mix up and down 10 times.
l Place the barrette in a magnetic rack for 2–3 min until the
supernatant is clear.
l Remove and discard the supernatant.
l Repeat the binding buffer washing step for a total of three
washes.
l Resuspend the beads in 200 μL of binding buffer.
Library capture:
l When the hybridization program is finished, place sample at RT.
l Transfer each hybridization reaction to the washed streptavidin
beads (see Note 21).
l Incubate for 30 min at 1800 rpm at RT.
l While the sample is incubated, prepare 200 μL 3 of wash
buffer for each sample in strip of PCR tubes. Incubate them at
65 C in the thermocycler.
l At the end of the 30 min, quickly spin down the samples, and
place them on a magnetic rack.
l When the supernatant is clear (~2 min), take it off, and add
200 μL of SureSelect Wash Buffer 1 on the beads.
l Mix up and down and close samples with new lids. Vortex for 5 s
the suspended beads.
l Quickly spin down and replace samples on the magnetic rack.
l When the supernatant is clear (~2 min), take it off.
l Place sample on a simple rack, and add 200 μL of SureSelect
Wash Buffer 2 at 65 C.
l Mix up and down 10 times (do not vortex); check the bead
suspension.
l Incubate tubes for 5 min at 65 C on the thermocycler.
l Spin down, and put back the sample on the magnetic rack, and
when the supernatant is clear, take it out.
l Change the lid between each wash.
l Repeat the wash with the SureSelect Wash Buffer 2 for a total of
three washes.
l Discard the supernatant of the last wash, and immediately resus-
pend the beads in 23 μL of ultrapure water (do not let the beads
dry).
l Mix up and down and keep selected library on ice.
Library Amplification Select the adapters P7 and P5 to add to each sample. Each sample
will have a unique combination of P5/P7 adapters. Always use one
of these P5 pairs: P5 i13 and P5 i14 or P5 i15 and P5 i16 or P5 i17
and P5 i18. When 12 or more different samples are sequenced in
the same run, use at least 1 different P5 in addition to the selected
pair.
l Prepare the PCR Mix in pre-PCR zone under hoods (Table 10)
by adding for each reaction 13.5 μL ultrapure water, 10 μL 5
Herculase II Rxn Buffer, 0.5 μL 100 nM dNTP Mix, and 1 μL
Herculase II fusion DNA Polymerase.
l Vortex and spin down.
l Add 25 μL of PCR Mix into each selected library sample.
l Mix up and down.
l Add 1 μL of selected primers P7 in the corresponding sample.
l Add 1 μL of selected primers P5 in the corresponding sample.
l Mix up and down.
Table 10
Reagents and volumes per reaction for PCR mix
Reagents Volume for 1 reaction (μL)

Ultrapure water 13.5
5 Herculase II Rxn Buffer 10
100 nM dNTP Mix 0.5
Herculase II fusion DNA Polymerase 1
Total volume 25
Table 11
Thermocycler program for library amplification

1 98 2 min
12 98 30 s
58 30 s
72 1 min
72 5 min
1 72 Hold
l Run the following program in PCR room (Table 11): 2 min at

98 C, 12 cycles (30 s at 98 C, 30 s at 58 C, and 1 min at
72 C), and 5 min at 72 C, and hold at 4 C.
l Quickly spin down the samples.
l Place samples on a magnetic rack.
l When the supernatant is clear, remove and transfer them in a
new strip of PCR tubes (see Note 16).
Purification with AMPure l Take out the beads 30 min before the experiment and leave them
Beads at RT. Vortex roughly to mix the beads until the reagent is
homogeneous.
l Add 60 μL of beads to each sample containing around 50 μL.
Mix up and down.
l Place sample on the magnetic rack and wait for 2 min. The
supernatant has to be clear.
without touching the beads.
l Incubate for 1 min and then take the ethanol out.
l Centrifuge the sample and replace them on the magnetic rack.
l Let the sample dry on the magnetic rack around 2 min at RT (see
Note 15).
l Mix the beads in 25 μL of nuclease-free water. Mix up and down.
l Incubate for 2 min at RT and then centrifuge.
l Take out the supernatant (around 25 μL) in a new 1.5 mL tube

(see Note 16).
l Name the tubes. Tubes can be stored at 4 C or 20 C.
Final Library Verification Dilute 1 μL of your library in a new tube at 1/3 with ultrapure
water. Use 1 μL of this dilution to load a DNA HS chip for
Bioanalyzer.
l Adjust the plate base of the chip priming station in position C
(Fig. 12).
l Adjust the syringe clip to the lowest position (Fig. 13).
l Incubate the kit at RT for 30 min before using it.
Prepare the gel-dye mix:
l Vortex the DNA dye concentrate (blue lid) for 10 s and
spin down.
l Add 15 μL of DNA dye concentrate into a DNA gel matrix tube
(red lid).
l Vortex the mix and transfer it in a spin filter.
l Centrifuge the spin filter at 2240 g for 10 min at RT.
l Discard the filter and label the tube with the preparation date.
The gel can be used for 4 weeks and stored in the dark at 4 C.
Loading the gel-dye mix:
l Place new DNA1000 chip in the priming station.
l Pipette 9 μL of gel-dye mix at the bottom of the well (Fig. 18).
l Set the timer to 60 s.
Fig. 18 Picture of HS DNA chip indicating where to load (well 15) the 9 μL of
gel-dye before priming the chip
Fig. 19 Picture of HS DNA chip indicating where to load (wells 13, 14, and 16) the
9 μL of gel-dye after priming the chip
l Place the plunger at 1 mL and then close the chip priming

station.
l Be sure that the lock clicks correctly.
l Press the plunger of the syringe down until it is held by the clip.
l Wait 60 s and release the plunger.
l Let the plunger move back until it stops and then slowly pull
back the plunger to the 1 mL position.
l Open the priming station.
l Add 9 μL of gel-dye mix in the other well-marked (Fig. 19).
Loading the chip:
l Add 5 μL of DNA marker (green lid) into all the wells you did
not add gel mix (ladder well included) (Fig. 20).
l Pipette 1 μL of ladder (yellow lid) in the well with a ladder. Add
1 μL of your sample per well or marker in unused well (Fig. 20).
l Vortex for 60 s at 2400 rpm in the vortex mixer IKA-Model
MS3.
l Load the chip within 5 min in the Bioanalyzer.
Start the run:
l Select the appropriate assay, click “dsDNA” and then “High
Sensitivity DNA.xsy” (Fig. 17).
l Indicate the run data destination.
l Enter you sample name (see Note 19).
Fig. 20 Picture of HS DNA chip for loading of marker, ladder, and samples. Load
5 μL of marker in well 1 to 12. Then load 1 μL of ladder in the well 12. Load 1 μL
of sample in wells 1 to 11 (one sample per well)
Electrode cleaning:
l Wash the electrodes, before and after every run, with the wash
chip loaded with 350 μL of deionized analysis grade water.
3.6.3 High-Throughput The day before remove the reagent cartridge and the hybridization
Sequencing buffer HT1 from 20 to 4 C overnight. Otherwise, thaw reagents
in RT water bath (~1 h) without submerging the cartridge. Gently
tap on the bench to dislodge water from the base and then dry
the base.
Library Preparation for l Dilute all libraries to 4 nM, 2 nM, 1 nM, or 0.5 nM with
Sequencing ultrapure water.
l Pool all libraries in a single tube by combining the same volume
of the 4 nM dilution.
l Take out the flow cell from the fridge at least 30 min before
the run.
l Prepare 1 mL of NaOH 0.2 N using 980 μL ultrapure water and
20 μL NaOH 10 N; mix the tube by inverting.
l Prepare 1 mL of 200 mM Tris–HCl, pH 7 using 800 μL ultra-
pure water + 200 μL Tris–HCl, pH 7 1 M.
l Clean the positions 7, 8, 9, 20, 21, and 22 of the reagent
cartridge; then pierce them.
l For the read 1, combine 2.7 μL SureSelect QXT Read Primer
1 with 897.5 μL of BP10 (Table 14).
Table 12
NextSeq 550 Mid-Output v2 Kit
Total Final
Sequencing Volume of volume cartridge
read Volume of SureSelect QXT Primer Illumina primer (mL) position
Read 1 2.7 μL SureSelect QXT Read Primer 1 (brown 897.3 μL BP10 0.9 Well 7
cap) (from well 20)
Read 2 3.3 μL SureSelect QXT Read Primer 2 (black 1096.7 μL BP11 1.1 Well 8
cap) (from well 21)
Index + index 4.8 μL SureSelect QXT Index Read Primer 1590.4 μL BP14 1.6 Well 9
2 (clear cap) + 4.8 μL SureSelect QXT Index (from well 22)
2 read Primer NSQ (purple cap)
Table 13
Volume of library and 0.2 N NaOH to use depending of the starting library
concentration
Starting library concentration (nM) Library (μL) 0.2 N NaOH (μL)

4 5 5
2 10 10
1 20 20
0.5 40 40
Table 14
Volume of prechilled HT1 to add for a library dilution at 20 pM
Starting library concentration (nM) Prechilled HT1 (μL)

4 985
2 970
1 940
0.5 880
l For the read 2, combine 3.3 μL SureSelect QXT Read Primer

1 with 1096.7 μL of BP11 (Table 14).
l For the Index and Index 2, combine 4.8 μL SureSelect QXT
Index Read Primer, 4.8 μL SureSelect QXT Index 2 Read Primer
NSQ, and 1590.4 μL of BP14 (Table 12).
l Add, respectively, Read 1, Read 2, and Index 1 and Index 2 to
positions 7, 8, and 9 (Table 14).
l Depending on the initial concentration of your library pool, mix

it with the NaOH solution. For a library pool concentration of
4 nM, mix 5 μL of library with the same volume of 0.2 N NaOH.
Proceed in the same way for 2 nM, 1 nM, and 0.5 nM of library
concentration with a volume of 10 μL, 20 μL, and 40 μL,
respectively (Table 13).
l Vortex and briefly centrifuge the mix.
l Incubate at RT for 5 min.
l Always depending on the initial concentration of your library
pool, mix it with 200 mM Tris–HCl, pH 7. For a library pool
concentration of 4 nM, 2 nM, 1 nM, and 0.5 nM, add 5 μL,
10 μL, 20 μL, and 40 μL, respectively, of Tris–HCl, pH 7.
l Dilute the library to set the concentration to 20 pM with HT1
(Table 14).
l Vortex and briefly centrifuge the mix. Place it on a
refrigerated rack.
l Final dilution for the library, usually between 1.2 and 1.4 pM for
QXT, is empirically adjusted depending on previous run results.
l According to the required concentration (1.1–2.0 pM), com-
bine the 20 pM library with HT1 buffer in a final volume of
1.3 mL (Table 15).
l Mix by inversion, briefly centrifuge, and place it on a
refrigerated rack.
Table 15
Dilution of the 20 pM library with HT1 buffer depending on the required
final concentration
Final required 20 pM library HT1 buffer

concentration (pM) volume (μL) volume (μL)
1.1 71.5 1228.5
1.2 78 1222
1.3 84.5 1215.5
1.4 91 1209
1.5 97.5 1202.5
1.6 104 1196
1.7 110.5 1189.5
1.8 117 1183
1.9 123.5 1176.5
2.0 130 1170
Start the Sequencing Run Preparation of the reagent cartridge and the flow cell
l Invert cartridge to mix reagent.
l Check if positions 29, 30, 31, and 32 are thawed. Gently tape on
the bench to reduce air bubble. Leave the flow cell for 30 min at
RT outside of its box. Clean the glass surface with a lint-free
alcohol wipe. Dry the glass with a low-lint lab tissue.
l Ensure that the flow cell ports are not obstructed and their joints
are well fixed. The white plastic tenons have to be visible. Check
that the four white pliers are triggered on the black support
plaque. The four metallic pliers are positioned lying down the
black support plaque (Fig. 21).
l Remove the new flow cell from its storage (2–8 C).
l Let the unwrapped flow cell at RT for 30 min. Avoid repeated
temperature variations.
l Remove the flow cell from the foil package.
l Open the clear plastic clamshell.
l Clean the glass surface with a lint-free alcohol wipe.
l Dry the glass with a low-lint lab tissue.
l Clean the foil seal covering reservoir #10 labeled “Load Library
here” using a low-lint tissue.
l Pierce the seal with a clean 1 mL pipette tip.
l Load 1.3 mL of prepared libraries into reservoir #10 labeled
“Load Library here.” Avoid touching the foil seal as you dis-
pense the libraries (Fig. 22).
l From the Home screen, select “Experiment,” and then select
“Sequence.” The sequence command opens the imaging
Fig. 21 Picture of the back of the mid-output flow cell

Fig. 22 Picture of the reagent cartridge
Fig. 23 Removing of the waste container
compartment door, releases consumables from a previous run,

and opens the series of run setup screens. A short delay is
normal.
l Remove the previous flow cell.
l Align the flow cell over the alignment pins, and place the flow
cell on the stage.
l Select “Load.”
l Click “Next.”
l Empty the waste container in an appropriate toxic liquid recy-
cling bin (contains formamide).
Fig. 24 Loading of the buffer cartridge
Fig. 25 Removable position #6 (a) with its protective rubber cover (b)
l Replace the waste container sliding it until an audible click

indicates that the container is in position (Fig. 23).
l Remove the used buffer cartridge, and slide a new one until an
audible click indicates that the cartridge is in position (Fig. 24).
l Close the buffer compartment door and click “Next.”
l Remove the used reagent cartridge. The set contains formamide
and needs to be treated as toxic reagents. To facilitate safe
disposal of unused reagent, remove the protective rubber cover
over the slot next to position #6. Press down on the clear plastic
tab, and push toward the left to eject the reservoir (Fig. 25).
l Slide the new reagent cartridge until it stops and loses the
reagent compartment door (Fig. 26).
Fig. 26 Loading of the new reagent cartridge
l Click “Load.”
l Select “Next” after the 30 s of cartridge loading.
l Run parameters
l Enter a run name of your preference.
l Enter a library ID of your preference.
l From the Recipe drop-down list, select a recipe. Only compati-
ble recipes are listed.
l Select the read type “Paired End.”
l Enter the number of cycles for each read in the sequencing run
(2*151) (see Note 22).
l Enter the number of cycles required for the Indices 1 and 2.
l Select “Next.”
l The software performs an automated check of the system.
l When the check is finished, select “Start.”
3.7 Identity As many steps are manual, it is crucial to ensure that final results
Monitoring correspond to the correct patient. A specific identity monitoring
should be implemented: panel should be designed with at least five
control SNV (single-nucleotide variant) (see Note 23) to have a correct
discrimination power. The predesign TaqMan PCR assays should be
performed independently on the extracted DNA. SNV genotype
would be also extracted HTS data: when both techniques give the
same results for five control probes (or 4 + SRY), DNA identities are
Table 16
Set up PCR mix for TaqMan reaction
Reagents Final mix Volume for 1 reaction (μL)

DNA (2 μg/mL) 5 ng 4.5
TaqMan primers (20) 1 0.5
TaqMan Master Mix (2) 1 5
Total volume 10
confirmed with a resolving power of 99% (1% of residual risk to

randomly observe the same genotypes from two distinct DNAs).
TaqMan PCR:
Probes (normal and mutated) are ordered either specifically
(Custom TaqMan Assay) or in a predesign kit (predesign TaqMan
Assay). One of the probes is tagged with FAM and the second one
with VIC.
In addition to the patient’s samples, add a negative control
(blank) and two control samples with a known genotype. The
plate can be prepared 24 h in advance and store at 4 C, but reading
must be done just after the PCR if the thermocycler does not
automatically read fluorescence.
l Set up a PCR mix for each SNP to check with 0.5 μL of TaqMan
probes (20) and 5 μL of TaqMan Master Mix 2 per sample to
test (plus one for negative control with ultrapure water).
l Add 4.5 μL of DNA at 2 μg/mL in one of each different SNP
mix (Table 16).
l Seal the plate and spin it before loading.
l Push the right button on the LightCycler® 480.
l Load the plate.
l Push the right button again to close.
l Open the software LightCycler® 480 SW 1.5.
l Click “New Experiment” (Fig. 27).
l Select a program where samples incubate for 10 min at 95 C,
50 cycles of 15 s at 95 C and then 1 min at 60 C, 1min at 37 C
(Table 17).
l Click “Strat Run.”
l Save the plate.
l Click “Subset Editor” and define a subgroup per SNP.
l Assign the SNP to their place in the plate.
l Click “Sample Editor” and select the workflow “Endpt Geno.”
Fig. 27 Home page of the software LightCycler® 480 SW 1.5 (Idaho Technology, Inc., Roche)
Table 17
PCR program for TaqMan PCR
Cycle number Target ( C) Acquisition mode Hold Ramp rate ( C/s)

1 95 None 10 min 4.4
50 95 None 15 s 4.4
60 Single 1 min 2.2
1 37 Single 1 min
l Enter sample name.

l At the end of the program, click “Analysis.”
l Select the subgroup (SNP) to analyze.
l Click “Calculate.”
l The genotype of each sample for each SNP is determined by the
fluorescence value obtained by the two probes.
l Compare these results with the SNP genotype extracted from

the HTS variant table.
l The identity is established if the genotypes match.
3.8 Results Analysis The bioinformatics pipeline implemented for the analysis of the
NGS data and named STARK (Stellar Tools from raw sequencing
3.8.1 Bioinformatics and
data Analysis to variant RanKing) is based on the GATK recom-
Software
mendations. The raw data are first processed by the NextSeq
550 (Illumina) integrated software named Real-Time Analysis
(RTA). The RTA software analyzes the images and attributes
bases (base calling) to produce a .bcl file. These .bcl files are con-
verted into sequences (reads) stored into .fastq files thanks to the
BCL2FASTQ software (also by Illumina). The reads demultiplex-
ing and the trimming of adaptors are performed during this step.
We recommend using quality analysis software like FASTQC to
check for quality metrics for each read (Q30). The alignment to
the reference human genome (GRCh37) is computed using the
BWA-MEM software that will produce the .bam files. The variant
calling is performed using both GATK HaplotypeCaller and GATK
UnifiedGenotyper to produce a combined .vcf files containing all
variations identified. Finally, the variant annotations are performed
by Alamut Batch (Interactive Biosoftware, Rouen, France), and
variant prioritization is performed by VaRank [3]. The output
files (tab-separated values) can be easily analyzed using a spread-
sheet program such as Microsoft Excel.
3.8.2 Variants Analysis The next objective is to find if there are any pathogenic variants in
this list. We advise to proceed sequentially with different filters
which are different for every panel. Here we describe the sequential
procedure used in the GenoDENT project:
l If you have access to the parent sequencing, search for de novo
variants.
l Search for CNV with the CANOES software [4].
l Analyze the pathogenic reported variants using DECIPHER,
HGMDPro, and ClinVar.
l Search for chromosome X variants filtering the ExAC MAF and
1000genome MAF at <5%.
l Search for autosomal recessive variants (homozygous or com-
pound heterozygous) filtering the ExAC MAF and
1000genome MAF at <5%.
l Search for autosomal dominant variants filtering the ExAC MAF
and 1000genome MAF at <1%.
l Once a pathogenic candidate is found, report all the arguments’

pros and cons of its pathogenicity, and assign it a class [5]:
pathogenic, likely pathogenic, variant of unknown significance,
likely benign, and benign.
4 Notes
1. To reduce the contamination risk, the laboratory should have

four different areas: the DNA extraction area, the PCR mix
pre-PCR activity area, the intermediate area, and the post-PCR
activity area. All these areas are separated and have their own
material.
2. Organize your genes in order to rank them easily following
different criteria (relevant to disease, project importance, gene
size, number of probes needed to sequence, etc.). This ranking
facilitates the selection of genes and saves time. Gene lists are
not exhaustive and will vary according to the scientific purpose
and monitoring and analysis of the new patients’ cohort.
3. The gene panel should contain at least four autosomal infor-
mative SNV and one SRY probe or five autosomal informative
SNV. In our case, an informative SNV has MAF (minor allele
frequency) close to 50%. They allow setting up an identity
control at the end of the workflow to ensure that the final vcf
results correspond to the same extracted DNA. This monitor-
ing is performed using a TaqMan assay detailed in 3.7 identity
monitoring.
4. Agilent has a pricing tier policy according to the total probe
number and the total capture size. If the design is borderline
one tier, it can be wise to modulate parameters in order to stay
in the cheapest tier. For example, it is possible to change the
boosting parameter to “balanced” for genes that are less criti-
cal, to decrease extension in intronic regions to 10 bp. Ulti-
mately reduce the number of tested genes.
5. In the design documents that can be downloaded, there are
details about all the regions that have been submitted. It is
important to look at these files as the software won’t be able to
design probes for some regions. If some specific regions are
necessary for the application, verify their coverage in the file
“Genomic regions expected to be sequenced” list. If some
regions are not covered, copy paste their genomic region
from this file to SureDesign, and resubmit these regions using
different parameters (e.g., by using the “No masking”
parameter).
6. Pay particular attention to the container and its flotation. The

container should not tip over, and make sure the written
patient identification does not vanish in the water bath.
7. Make sure to let sample at least 2 min at RT, and then measure
them with the Qubit.
8. We recommend adding at each run a DNA sample of a known
concentration to control the calibration curve (IQC, internal
quality control). Measure it before and after the samples.
9. Use calibrated pipette for this quantification. We do not rec-
ommend pipetting 1 μL of sample due to the possible pipetting
errors.
10. When you move the chip outside its box, pay attention to the
delicate zipper.
11. If the chip is taken out, the priming step will be done again.
You can use the priming time to prepare your sample plate.
12. The preheat lasts between 5 and 10 min, and the migration
takes 3 min per sample.
13. It is a critical step, because too much DNA will decrease
enzyme efficiency.
14. Be careful and pipette slowly because the supernatant is
viscous.
15. Do not let the beads crack; when they are of matt appearance, it
is most likely that they are dry.
16. Keep the bead for safety.
17. Stop point: Samples can be stored at 4 C for 1 week or 20 C
for a longer preservation.
18. The Bioanalyzer is sensible to any vibrations, which can alter
the results.
19. Fill sample name anytime.
20. If possible, use the maximum quantity of DNA for the
hybridization.
21. Hybridization volume should be around 30 μL, but not under
25.5 μL.
22. Always add one cycle to the wanted number for the phasing/
prephasing.
23. In addition to the sex identification, here are few examples of
SNV (Table 18): rs2292597 (PRSS12), rs529208 (DOCK8),
rs3735803 (TRAPPC9), rs2306331 (AP4E1), rs7301328
(GRIN2B), and rs1047179 (FTCD).
Table 18
Example of SNP for identity monitoring
N Gene N rs
SNV1 PRSS12 rs2292597
SNV2 DOCK8 rs529208
SNV3 TRAPPC9 rs3735803
SNV4 AP4E1 rs2306331
SNV5 GRIN2B rs7301328
SNV6 FTCD rs1047179
Acknowledgments
We are grateful to all patients and their families for their participation
and invaluable contribution. We would like to thank also Dr Karen
Niederrheiter for critical reading of the manuscript. We would like to
acknowledge colleagues from the O-Rares French Network for refer-
ring patients: Pr Marie-Cécile Maniere, Pr François Clauss, Dr Fré-
déric Obry, Dr Sophie Jung, Dr Marion Strub, Dr Bruno
Grollemund, Pr Olivier Huck, Pr Maryline Minoux, Pr Béatrice
Walter, Dr Olivier Etienne, Dr Etienne Waltman, Dr Catherine
Gros, Dr Fabien Bornert, and Dr Sophie Bailly, CRMR Coordinat-
ing Center, Hôpitaux Universitaires de Strasbourg, Pôle de Méde-
cine et Chirurgie Bucco-Dentaires; Pr Ariane Berdal, Dr Muriel de la
Dure Molla, and Dr Benjamin Fournier, Constitutif Center, Hôpital
Rothschild, Paris, Odontologie; Dr Victorin Ahossi, CCMR, CHU
de Dijon, Odontologie; Pr Marie-José Boileau, CCMR, CHU de
Bordeaux Hôpital Pellegrin, Odontologie et Santé Buccale; Dr Ser-
ena Lopez Cazaux, CCMR, CHU de Nantes Hôtel Dieu, Odonto-
logie Conservatrice et Pédiatrique; Pr Frédéric Cuisinier, CCMR,
CHU de Montpellier Hôpital Gui de Chauliac, Odontologie; Pr
Vianney Descroix, CCMR, Groupe Hospitalier Pitié-Salpêtrière
Paris, Odontologie; Dr Frédérique Dhalluin-Olive, CCMR, Centre
Hospitalier d’Angoulème, Odontologie; Dr Edouard Euvrard,
CCMR, CHU de Besançon Hôpital Jean Minjoz, Chirurgie
Maxillo-Faciale, Stomatologie et Odontologie; Dr Marie-Paule
Gelle, CCMR, CHU de Reims Hôpital de la Maison Blanche,
Odontologie; Pr Bruno Gogly, CCMR, CHU Henri Mondor
Paris, Odontologie; Dr Hervé Moizan, CCMR, CHU de Rouen
Hôpital Saint Julien, Stomatologie; Pr Jean-Jacques Morrier and Dr
Jean-Jacques Duprez, CCMR, Hospices Civils de Lyon Groupement
Hospitalier Centre, Consultations et Traitements Dentaires; Dr
Dominique Droz, CCMR, CHU de Nancy Hôpital Brabois,
Odontologie; Pr Jean-Louis Sixou and Pr Martine Bonnaure-Mallet,

CCMR, CHU de Rennes, Odontologie et Chirurgie Buccale; Dr
Florent Sury, CCMR, CHU de Tours Hôpital Trousseau, Chirurgie
Maxillo-Faciale; Pr Corinne Tardieu, CCMR, Hôpital La Timone
Marseille, Odontologie, Pr Michèle Muller-Bolla, Nice; and Pr Fré-
déric Vaysse, Pr Isabelle Bailleul Forestier, CCMR, CHU de Tou-
louse Hôpital Rangueil, Odontologie.
The authors declare no potential conflict of interest with
respect to the authorship and/or publication of this article.
Funding
This work was supported by grants from the French Ministry of
Health (National Program for Clinical Research, PHRC
2008 N 4266 Amelogenesis imperfecta), the University Hospital
of Strasbourg (API, 2009–2012, “Development of the oral cavity:
from gene to clinical phenotype in patients”), and the EU-funded
projects (ERDF) A27 “Orodental manifestations of rare diseases,”
supported by the RMT-TMO Offensive Sciences initiative, INTER-
REG IV Upper Rhine program (www.genosmile.eu) and RARE-
NET (www.rarenet.eu) INTERREG V Upper Rhine program.
References
1. Bloch-Zupan A, Sedano H, Scully C (2012) 4. Backenroth D, Homsy J, Murillo LR et al
Dento/oro/craniofacial anomalies and genetics, (2014) CANOES: detecting rare copy number
1st edn. Elsevier, Boston, MA variants from whole exome sequencing data.
2. Prasad MK, Geoffroy V, Vicaire S et al (2016) A Nucleic Acids Res 42(12):e97
targeted next-generation sequencing assay for 5. Richards S, Aziz N, Bale S et al (2015) Standards
the molecular diagnosis of genetic disorders and Guidelines for the Interpretation of
with orodental involvement. J Med Genet Sequence Variants: A Joint Consensus Recom-
53:98–110 mendation of the American College of Medical
3. Geoffroy V, Pizot C, Redin C, Piton A, Vasli N, Genetics and Genomics and the Association for
Stoetzel C, Blavier A, Laporte J, Muller J (2015) Molecular Pathology. Genet Med 17
VaRank: a simple and powerful tool for ranking (5):405–424
genetic variants. PeerJ 3:e796. https://doi.org/
10.7717/peerj.796
Annex 1
GenodentV04_0 513 genes
AAAS AXIN2 CLDN15 DLL1 FERMT3 GTF2I KAT6B LTF OBSL1 PROKR2 SLC10A7 SUOX TUFT1
ABCA5 B3GAT3 CLDN16 DLX3 FGD1 GTF2IRD1 KATNB1 LYSMD4 OCRL PSAP SLC13A5 TACR3 TWIST1
ACPT B4GALT7 CLDN19 DLX5 FGF10 H19 KAZN LYST ODAM PTCH1 SLC20A1 TBCE TWIST2
ACVR1 BANF1 CLEC7A DMP1 FGF23 H1FNT KCNH1 MAP2K1 OFD1 PTCH2 SLC20A2 TBX1 UBB
ADAM10 BAZ1B CLRN1 DNM1 FGF3 HCCS KCNJ2 MAP2K2 ORAI1 PTDSS1 SLC24A4 TBX2 UBE3B
ADAM15 BCOR CNNM4 DOCK8 FGF8 HEMK1 KCTD1 MASP1 ORC1 PTH1R SLC26A2 TBX22 UBR1
ADAMTS1 BLM COG6 DOK2 FGF9 HENMT1 KDM6A MBTPS2 OSTM1 PTHLH SLC29A3 TBX3 UHRF1
ADAMTS10 BMP1 COL10A1 DSG4 FGFR1 HMCN1 KIF7 MED12 PAX3 PTPN11 SLC34A1 TCEAL7 USH1C
ADAMTS2 BMP2 COL11A1 DSP FGFR2 HNF1B KISS1 MED25 PAX9 PVRL1 SLC34A2 TCIRG1 USH1G
ADGRV1 BMP4 COL17A1 DSPP FGFR3 HOXB13 KISS1R MEGF8 PCDH15 PVRL4 SLC34A3 TCOF1 USH2A
ADNP BRAF COL1A1 DVL1 FKBP10 HOXD13 KL MEPE PCNT RAB23 SLC35C1 TCTEX1D2 USP19
AHCY C1R COL1A2 DVL3 FLNA HRAS KLK4 MID1 PDGFRB RAI1 SLC39A13 TERC VAV1
AIP C1S COL24A1 EDA FLNB HSPG2 KMT2D MMP1 PDZD7 RAPSN SLC9A3R1 TERT VDR
AIRE C4ORF26 COL2A1 EDAR FN1 IBSP KRAS MMP14 PEX1 RASGRP2 SMAD3 TFAP2A VIPAS39
AKAP13 CA2 COL3A1 EDA FOXC1 IDS KREMEN1 MMP2 PEX6 RBBP8 SMARCAL1 TFAP2B VPS13A
RADD
AKT1 CAMK4 COL5A1 EFNB1 FRAS1 IDUA KRT14 MMP20 PHEX RBM28 SMOC2 TGFA VPS33B
ALDH3A2 CARD9 COL5A2 EHMT1 FREM2 IFIH1 KRT16 MMP9 PIGA RECQL4 SNRPN TGFB1 VPS4B
ALPL CASP14 COL7A1 EIF2AK1 GALC IFITM5 KRT17 MPPE1 PIGL RFC2 SNX33 TGFB2 WDR19
ALX3 CCBE1 COL9A1 ELN GALNS IFT122 KRT5 MSX1 PIK3R1 RIN2 SOS1 TGFB3 WDR35
AMBN CCDC8 COL9A2 EMR2 GALNT3 IFT140 KRT6A MSX2 PITX2 RMRP SOST TGFBR1 WDR6
AMELX CD96 COX7B ENAM GAS1 IFT20 KRT6B MTERF2 PKP1 RNF10 SOX10 TGFBR2 WDR72
Protocol GenoDENT
AMELY CDC6 CREBBP ENPP1 GBP5 IFT43 KRT6C MUTYH PLEC ROGDI SOX11 TGIF1 WDR83
AMER1 CDH1 CROT EP300 GDF5 IGF1R KRT83 MYCBP2 PLEKHM1 ROR2 SOX18 THRA WHSC1
451
(continued)
452
AMTN CDH23 CRTAP ERCC3 GJA1 IGSF3 LAMA3 MYO1H PLG RPS6KA3 SOX2 TINF2 WISP3
ANKH CDH3 CRTC2 ERCC4 GJB3 IKBKG LAMB3 MYO7A PLK4 RUNDC1 SOX21 TMCO1 WNT1
Tristan Rey et al.
ANKRD11 CDKN1C CSNK1A1 ERCC8 GJB4 IL11RA LAMC2 NAA10 PLOD1 RUNX2 SP6 TMEM38B WNT10A
ANTXR1 CDON CSPP1 EVC GJB6 IL17F LEF1 NACC1 PLXNB2 SALL4 SP7 TNFRSF11A WNT10B
ANTXR2 CELSR3 CTNNB1 EVC2 GLA IL17RA LEMD3 NDN PLXNB3 SAMD12 SPARC TNFRSF11B WNT3
AP1G2 CENPJ CTSC EXT1 GLB1 INSR LEPRE1 NFKBIA PLXND1 SAT1 SPARCL1 TNFSF11 WNT5A
AP3B1 CEP152 CTSK EXT2 GLI2 IPO4 LIFR NHS POC1A SATB2 SPECC1L TP63 WRN
APAF1 CFDP1 CUL7 EYA1 GLI3 IRF6 LIMK1 NIPBL POLD1 SCARF2 SPP1 TRAF6 XPR1
APC CHD7 CYP27B1 EZR GNAS IRX5 LMNA NKG7 POLR1C SEC23A SPRY4 TRIM37 ZEB1
AR CHPF DCAF17 FADD GOLIM4 ITGA6 LONP1 NOP10 POLR1D SEC24D SQSTM1 TRIP10 ZEB2
ARHGAP6 CHST3 DEPTOR FAM111B GORAB ITGB2 LRFN2 NOTCH2 POLR3A SERPINF1 SSUH2 TRIP11 ZFHX4
ARID1B CHSY1 DFNB31 FAM20A GPC3 ITGB4 LRP4 NOTCH3 POLR3B SERPINH1 STAT1 TRPS1 ZFPM1
ATP6V0A2 CIAPIN1 DGKG FAM20B GPR68 ITGB6 LRP5 NRAS PORCN SH3BP2 STAT3 TRPV3 ZIC2
ATP6V1B2 CIB2 DHCR24 FAM20C GREM1 ITPR3 LRP6 NSD1 PPIB SH3PXD2B STIM1 TSC1 ZMPSTE24
ATR CKAP2L DHCR7 FAM83H GREM2 JAG1 LTBP2 NSUN2 PRKAR1A SHH SUFU TSC2 ZNF469
ATRIP CLCN7 DHODH FBN1 GRHL2 KAL1 LTBP3 NTRK1 PROK2 SIX3 SUMO1 TSPEAR ZNF878
ATRX CLDN1 DKC1 FERMT1 GRHL3 KANSL1

Chapter 37
Protocols for Genetic and Epigenetic Studies of Rare

Diseases Affecting Dental Tissues
Bruna Rabelo Amorim, Pollyanna Almeida Costa dos Santos,
Caroline Lourenço de Lima, Denise Carleto Andia, Juliana Forte Mazzeu,
and Ana Carolina Acevedo
Abstract
This chapter describes methods related to the diagnosis of genetic dental diseases. Based on the present
knowledge, clinical phenotyping and next-generation sequencing techniques are discussed. Methods
necessary for Sanger sequencing, multiplex ligation-dependent probe amplification, and epigenetic modifi-
cation methods are detailed. In addition, protocols for cell culture establishment and characterization from
patients with inherited dental anomalies are described.
Key words Amelogenesis imperfecta, Tooth agenesis, Dentin inherited disorders, Next-generation
sequencing, Epigenetics, Cell culture
1 Introduction
Odontogenesis is a complex developmental process that involves

multiple and overlapping molecular events in signaling pathways
between the epithelium and cranial neural crest-derived ectome-
senchyme [1, 2]. Disturbances during tooth development may
result in dental anomalies. The etiology of dental anomalies is
mainly genetic, albeit epigenetic, local/systemic environmental,
and multifactorial basis also occur [3–5]. Dental anomalies can
manifest either as an isolated trait or as part of a syndrome. Broadly,
they can be classified into those affecting the teeth number, shape,
size and position, the structure of the mineralized tissues, and
dental eruption [6, 7]. Since 1996, an increasing effort has been
done in unraveling the molecular basis of several inherited dental
anomalies, and a significant progress has been made [8, 9]. Indeed,
numerous pathogenic variants associated with isolated and syndro-
mic forms of tooth malformations have been identified (https://
www.ncbi.nlm.nih.gov/omim). Among all the genetic dental
453
454 Bruna Rabelo Amorim et al.
disorders those regarding tooth agenesis, enamel and dentin for-

mation are the more extensively studied.
Tooth agenesis is the most common developmental craniofacial
abnormality in humans. It is defined as the absence of deciduous
and/or permanent teeth due to a development failure during
odontogenesis. Tooth agenesis can be classified as hypodontia,
oligodontia, or anodontia according to the number of teeth absent.
Hypodontia describes the absence of one to six teeth, whereas the
term oligodontia is applied to agenesis of more than six teeth
excluding third molars. The absence of all teeth is termed anodon-
tia and is a rare condition mainly associated with syndromes [10].
Tooth agenesis is present in syndromes such as ectodermal dyspla-
sias, Van der Woude syndrome (OMIM # 119300, 606713), oral–-
facial–digital syndrome type I (OMIM # 311200), and Rieger
syndrome (OMIM# 180500), among several others [11, 12].
The genetic basis of non-syndromic tooth agenesis, also
defined as selective tooth agenesis (STHAG), is supported by
molecular studies of non-syndromic familial autosomal dominant
tooth agenesis associated with mutations in genes expressed in early
tooth development. In recent years, X-linked inheritance patterns
have been reported [13]. At present, variants in nine genes have
been showed to cause selective non-syndromic tooth agenesis
(Table 1; insert here).
Amelogenesis imperfecta (AI) is a heterogeneous group of
genetic diseases that affects the quality and quantity of dental
enamel [17]. In both, non-syndromic and syndromic forms of AI,
Table 1
Genes involved in STHAG with well-characterized dental features
Ref
gene Locus Gene Phenotype Inheritance OMIM or ref.
STHAG
603507 12p13.2 LRP6 Oligodontia AD # 603507
142983 4p16.2 MSX1 Oligodontia/hypodontia AD # 106600
with or without
orofacial cleft
167416 14q13.3 PAX9 Oligodontia/hypodontia AD # 604625
604025 17q24.1 AXIN2 Oligodontia AD Lammi et al. (2004) [14];
Bergendal et al. (2011)
[15]
606268 2q35 WNT10A Oligodontia/hypodontia AD # 150400
300451 Xq13.1 EDA Oligodontia/hypodontia XLD # 313500
604095 2q13 EDAR Oligodontia/hypodontia AD Arte et al. (2013) [16]
606603 1q42-q43 EDARADD Oligodontia AD Bergendal et al. (2011)
[15]
608832 1q43 GREM2 Hypodontia AD # 617275
STHAG selective tooth agenesis
Molecular Diagnosis of Genetic Dental Anomalies 455
the deciduous and permanent dentitions are affected, although

with variable severity. The inheritance of non-syndromic AI can
be autosomal dominant (AD), autosomal recessive (AR), and
X-linked. AI can be classified according to the type of enamel
defect: (1) hypoplastic, with reduced enamel thickness, and
(2) hypomineralized, with normal thickness but impaired minerali-
zation. The hypomineralized AI can be also referred as hypocalci-
fied AI (enamel with reduced mineralization) and hypomaturation
AI (enamel with crystallite maturation defects). In some cases, the
classification of AI can be difficult since both hypoplastic and
hypomineralized phenotypes can be present in a family and even
in a same patient. To date, pathogenic variants in 16 genes have
been reported in non-syndromic AI, and pathogenic variants in
12 additional genes have been implicated in syndromic AI with
well-characterized dental phenotypes (Table 2; insert here).
Genetic diseases that primarily affect dentin formation are
known as heritable dentin structural diseases. They can be asso-
ciated with other symptoms like progressive sensorineural hearing
loss (OMIM#605594) or in syndromes such as osteogenesis
imperfecta and Goldblatt syndrome (OMIM#184260) [19–21].
To date, all reported dentin diseases exhibit an AD pattern of
inheritance and are classified into two major groups with further
subclassifications: dentin dysplasia (DD) with types I and II and
dentinogenesis imperfecta (DGI) with types I–III [22, 23]. More
recently, a more comprehensive classification of the non-syndromic
dentin disorders has been proposed and classifies the conditions in
two pathologies: (1) dentinogenesis imperfecta (formerly DGI
type I and II and DD type II) and dentin dysplasia (formerly DD
type I) [24]. The etiology of DGI is related to pathogenic variants
in one single gene DSPP (4q21.3), belonging to the SIBLINGs
family and encoding three major non-collagenous dentin matrix
proteins – dentin sialoprotein (DSP), dentin glycoprotein (DGP),
and dentin phosphoprotein (DPP) [25–27]. To date, 40 patho-
genic variants have been described in the DSPP gene (http://www.
lovd.nl/DSPP). Regarding the radicular DD, heterozygous path-
ogenic variants in two genes (VPS4B and SSUH2) were recently
reported [28, 29].
This increasing bulk of molecular evidence has an impact on the
better comprehension of tooth development, on the nosology of
genetic dental disorders, as well as in patients’ management and
family counseling.
The aim of the present chapter is to detail protocols per-
formed in the studies of genetic dental disorders, with a focus on
non-syndromic forms. For this purpose, this chapter is divided in
two parts. In the first part, based on the present knowledge, the
clinical phenotyping and choice of the sequencing methods are
discussed. Also, all methods necessary for Sanger sequencing are
detailed, as well as the multiplex ligation-dependent probe
Table 2
Genes involved in isolated and syndromic cases of AI with well-characterized dental features
Isolated AI
300391 Xp22.2 AMELX Hypoplastic/hypomaturation X-LINKED # 301200
606585 4p13.3 ENAM Hypoplastic AD/AR # 104500
# 204650
611927 8q24.3 FAM83H Hypomineralization AD # 130900
603767 19q13.41 KLK4 Hypomaturation AR # 204700
604629 11q22.2 MMP20 Hypomaturation AR # 612529
613214 15q21.3 72WDR72) Hypomaturation AR # 613211
614829 4q21.1 C4orf26 Hypomaturation AR # 614832
147558 2q24.2 ITGB6 Hypoplastic/hypomaturation AR # 616221
609840 14q32.12 SLC24A4 Hypomaturation AR # 615887
113811 10q24.3-25.1 COL17A1 Hypoplastic AD Prasad et al.
(2016) [9]
600805 18q11.2 LAMA3 Hypoplastic AD Prasad et al.
(2016) [9]
150310 1q32.2 LAMB3 Hypoplastic AD # 104530
601259 4q13.3 AMBN Hypoplastic AR # 616270
606362 19q13.33 ACPT Hypoplastic AR # 617297
610912 4q13.3 AMTN Hypomineralization AD Smith et al.
(2016) [18]
601404 14q32.11 GPR68 Hypomaturation AR # 617217
Syndromic AI
113811 10q24.3-25.1 COL17A1 AI/junctional epidermolysis AR/AD # 226650
bullosa
600525 17q21.33 DLX3 AI, hypomature–hypoplastic AD # 104510
with taurodontism
Tricho-dento-osseous syndrome # 190320
607805 2q11.2 CNNM4 Jalili syndrome AR # 217080
611062 17q24.2 FAM20A Enamel renal syndrome AR # 204690
614574 16p13.3 ROGD1 Kohlschütter–Tönz syndrome AR # 226750
611061 7p22.3 FAM20C Raine syndrome AR # 229775
600805 18q11.2 LAMA3 Junctional epidermolysis bullosa AR/AD # 226700
150310 1q32.2 LAMB3 Junctional epidermolysis bullosa AR/AD # 226700
# 226650
147557 17q25.1 ITGB4 Junctional epidermolysis bullosa AR/AD # 226650
# 226730
603959 3q28 CLDN16 Hypomagnesemia 3, renal AR # 248250
610036 1p34.2 CLDN19 Hypomagnesemia, renal with AR # 148290
ocular involvement
602090 11q13.1 LTBP3 Platyspondyly with amelogenesis AR # 601216
imperfecta
Adapted from Yamaguti et al. (2016)
amplification (MLPA) method. A description of the methodolo-

gies used to analyze epigenetic modifications is also discussed. In
the second part, protocols for cell culture establishment and char-
acterization from patients with inherited dental anomalies are
described.
2 Part 1
2.1 Phenotyping and The diagnosis of inherited dental anomalies needs a complete clini-
Sequencing cal assessment (physical examination, complementary imaging
Techniques exams, and histopathological analyses) that will allow for a careful
and detailed phenotyping. Environmental or systemic factors that
can induce a chronological development disturbance must be
excluded. In anamnesis, the family history of dental defects can be
indicative of a genetic condition, and the establishment of a likely
inheritance pattern should be performed. While in more severely
affected patients, the syndromic status can easily be recognized, a
detailed anamnesis can point to mild associated manifestations. A
syndromic case should be suspected whenever a dental anomaly is
associated to short stature, skeletal dysplasias, skin or hair defects,
heart defects, and systemic/metabolic conditions. Special attention
should be given to ectodermal defects in individuals with tooth
agenesis as well as to bone abnormalities in individuals with denti-
nogenesis imperfecta. Whenever a syndromic case is suspected, the
patient must be referred to a clinical geneticist. A thorough oral
evaluation can also aid the geneticist in the clinical diagnosis and
differential diagnosis of related conditions. This is the case of the
enamel renal syndrome (ERS, OMIM#204690) where the
oro-dental phenotype is pathognomonic and can be diagnosed
before the renal diagnosis [24].
Finally, the molecular analysis is necessary in order to define the
diagnosis. The family history, inheritance pattern, and comprehen-
sive phenotype will determine the choice of sequencing technique.
The strategy for selecting family members that will undergo
sequencing depends greatly in the inheritance patterns. When a
recessively inherited condition is suspected, the parents and siblings
must be tested. If a de novo variant is suspected, the biological
parents must be also tested.
2.2 Sequencing Genomic sequencing techniques have evolved rapidly since the
Techniques and the development of the first generation sequencing methods
Diagnosis of Inherited [30, 31]. Since then, different technologies for high throughput
Dental Diseases DNA sequencing have been developed. New-generation sequenc-
ing (NGS) techniques regroup platforms capable of generating
information on millions of base pairs in a single run. The different
NGS platforms available at present include 454-Roche, Illumina
Genome Analyzer-HiSeq/MiSeq and SOLiD, Ion Torrent and
PacBio RS, and Nanopore. Although the platforms differ consider-
ably from each other, all new-generation sequencings rely on mas-
sive parallel processing of DNA fragments. These sequencing
techniques have been applied for whole genome sequencing
(WGS), whole exome sequencing (WES), and targeted next-
generation sequencing panels. They have been very useful for the
diagnosis of Mendelian conditions [32–35].
It is important to note that NGS techniques are most useful

for the detection of single-nucleotide substitutions and insertions
or small deletions of 8–10 nucleotides or less [36]. Larger deletions
must be tested using other methodologies such as Multiplex
Ligation-dependent Probe Amplification (MLPA) [37, 38].
Another limitation of NGS is that high content of repeat sequences
present difficulties in assembly. However, single molecule real-time
sequencing (SMRT) has shown to be useful in sequencing through
extended repetitive regions in the genome [39].
The use of NGS has been particularly useful in the diagnosis of
genetic dental disorders. WES has enabled the diagnosis of syndro-
mic and non-syndromic AI [40–43]. A rapidly growing list of
disease-causing genes revealed the genetic heterogeneity of AI
and tooth agenesis, and a targeted resequencing strategy has been
proposed. In 2016, the first targeted NGS diagnostic panel for
genetic dental disorders was reported. The gene panel with
known disease-causing genes for isolated and syndromic amelogen-
esis imperfecta (AI), STHAG, and DGI, isolated DD, otodental
dysplasia, and primary failure of tooth eruption was reported [9].
At present, the use of targeted NGS can serve as a primary screening
tool before the application of WES/WGS. In some cases, such as
the ERS, where the clinical phenotype is pathognomonic, direct
Sanger sequencing of FAM20A gene is recommended. Regarding
DSPP associated dentinogenesis imperfecta, the single molecule
real-time (SMRT) DNA sequencing system has been proposed in
order to accurately characterize the human DPP repetitive region
without cloning [44] (Fig. 1; insert here).
Genetic dental disorders
Known genes Unknown genes
Single gene All WES

e.g. FAM20A known genes
(+) result
Sanger Targeted Validation

sequencing NGS Sanger sequencing
(-) result (-) result (+) result
MLPA Validation
del/dup Sanger sequencing
Fig. 1 Flowchart showing the sequencing techniques proposed for genetic dental disorders. MLPA Multiplex
Ligation Probe Amplification, ins insertion, dup duplication
As the protocols for NGS vary according to the platform and

equipment chosen, in the next section, only the methods preceding
Sanger sequencing (DNA isolation, amplification, and purification)
will be detailed. However, regardless of the sequencing technique
chosen, the first step is the DNA isolation.
3 Materials
3.1 DNA Isolation Two biological sources for DNA isolation are detailed, blood and
saliva [45] both by the salting out method [46] (see Note 1). The
DNA testing results from extracting DNA-rich cells by saliva or
blood samples are the same [47, 48]. The only differences are in
sample processing. The advantage of saliva sampling is that it is not
an invasive method. However, if the saliva sample is not dried and
stored appropriately, the risk of bacterial contamination is greater
than blood samples, as clean blood is collected in proper tubes,
which minimizes chances of contamination.
3.1.1 Materials Centrifuge tubes 15 mL.

200 and 1000 μL pipettes.
Pasteur pipettes.
Heater dry block or incubator.
Vortex.
Disposable conical polypropylene tubes (15 mL).
Polypropylene tube–microtube (2 mL).
3.1.2 Reagents (see 1. Blood Samples

Note 2) Ultrapure water.
RBC lysis solution: 5 mM MgCl2, 1 mM ethylenediaminete-
traacetic acid (EDTA), pH 8.0 (see Note 3).
Cell lysis solution: 10 mM Tris pH 7.5, 1 mM EDTA, pH 8.0,
1% sodium dodecyl sulfate (SDS) (see Note 4).
Protein precipitation solution: 7.5 M ammonium acetate
(NH4Ac) (see Note 5).
Isopropanol.
Ethanol.
Tris EDTA (TE) buffer 1: 10 mM Tris (pH 7.8) and 1 mM
EDTA (see Note 6).
2. Saliva Samples
Ultrapure water.
Sucrose 3%: weight 3 g of sucrose and add ultrapure water to a
volume of 100 mL.
TNE solution: 17 mM Tris–HCl (pH 8.0), 50 mM NaCl, and

7 mM EDTA diluted in ultrapure water (see Note 7).
TNE solution with ethanol: 17 mM Tris–HCl (pH 8.0),
50 mM NaCl, and 7 mM EDTA diluted in 66% ethanol
(see Note 8).
Lysis solution: 10 mM Tris (pH 8.0), 0.5% SDS, 5 mM EDTA
(see Note 9).
Proteinase K (Sigma Chemical Co., St. Louis, MO, USA)
(20 mg/mL). Store at 4 C.
Solution of EDTA and ammonium acetate (NH4Ac): 8 M
NH4Ac, 1 mM EDTA (see Note 10).
Isopropanol.
Ethanol.
TE buffer 1: 10 mM Tris (pH 7.8) and 1 mM EDTA (see
Note 6).
3.2 DNA Polymerase chain reaction (PCR) is a technology used for quick and
Amplification by PCR easy amplifying DNA sequences, which is based on the principle of
enzymatic replication of the nucleic acids. The PCR techniques are
used to sequence genes and can, thus, identify mutations in genetic
diseases. PCR facilitates cloning of DNA sequences and forms a
natural basis for cycle sequencing by the Sanger method. Here, a
basic PCR protocol is presented: the technique requires a repetitive
series of the three fundamental steps that defines one PCR cycle
(double-stranded DNA template denaturation, annealing of two
oligonucleotide primers to the single-stranded template, and enzy-
matic extension of the primers to produce copies that can serve as
templates in subsequent cycles [49, 50]).
3.2.1 Materials Thermocycler.

Polypropylene tube—microtube (0.2 mL).
3.2.2 Reagents Ultrapure water.

10 standard reaction buffer (see Note 11).
Magnesium chloride (MgCl2).
Deoxynucleotide triphosphates (dNTPs).
Forward primer.
Reverse primer.
Polymerase enzyme.
Target DNA.
3.3 PCR Products The PCR must be purified in order to eliminate products’, pri-
Purification mers’, and dNTPs’ contaminations that prevent sequencing. This
step is usually performed with appropriate kit (e.g., Qiagen, Pro-
mega, Affymetrix, Roche, etc.) (see Note 12).
3.3.1 Materials Microcentrifuge.

Polypropylene tube—microtube (1.5 or 2 mL).
3.3.2 Reagents (a) Purification by column

QIAquick® PCR Purification (Qiagen).
M sodium acetate, pH 5.0.
(b) Purification by agarose PCR gel product
Agarose.
TE buffer 1: 10 mM Tris (pH 7.8) and 1 mM EDTA (see
Note 6).
Ethidium bromide.
(c) Purification by ExoSAP-IT® (Affymetrix).
Exonuclease I and Shrimp Alkaline Phosphatase (ExoSAP-
IT®, Affymetrix).
4 Methods
4.1 DNA isolation Step 1: Cell Lysis

4.1.1 For Blood 1. Add 3 mL of blood to a disposable conical polypropylene tube
(15 mL) containing 9 mL of the RBC lysis solution.
2. Invert the tube and incubate at room temperature for 10 min.
Invert, at least, one more time during incubation.
3. Centrifuge for 10 min at 2000 g.
4. Remove the supernatant leaving a visible white cell pellet and
100–200 μL of residual liquid.
5. Vortex the tube vigorously to resuspend the white cells in the
residual supernatant.
6. Add 3 mL of the cell lysis solution to the tube containing the
resuspended cells and mix with transfer pipette (Pasteur) sev-
eral times until the solution becomes homogeneous. After
mixing, no cell residue (or cell pellet) should be visible. If
residues are present, incubate at 37 C until the solution
becomes homogeneous. The sample is stable if stored in lysis
solution at room temperature for 18 months.
Step 2: Protein Precipitation

1. Add 1 mL of the protein precipitation solution to the cell lysate
(room temperature).
2. Vortex vigorously for 20 s to mix the solution.
3. Centrifuge at 2000 g for 10 min. The precipitated proteins
will form a dark brown and compact pellet.
Step 3: DNA Precipitation
1. Transfer the supernatant into a 15 mL conic tube containing
3 mL of 100% isopranol.
2. Invert the tube slowly about 50 times until DNA “balls”
appear.
3. Centrifuge at 2000 g for 3 min. The DNA will be visible as a
small white pellet.
4. Remove the supernatant and drain the conic tube on absorbent
paper.
5. Add 3 mL of 70% ethanol. Invert the tube 20 times to wash the
DNA pellet.
6. Centrifuge at 2000 g for 1 min.
7. Remove the supernatant carefully. The pellet may be loose, so it
is necessary to invert the tube slowly and carefully.
8. Drain the conic tube onto absorbent paper and let the sample
to “dry” at room temperature for 15 minutes.
Step 4: DNA Hydration
1. Add 200–250 μL of 1 TE buffer (or water) which results in a
concentration of approximately 400 μg/μL.
2. Let the DNA to hydrate in this buffer at room temperature for
12–24 h or alternatively; incubate the DNA at 65 C for 1 h.
3. Proceed to downstream application, or store the DNA at
20 C.
4.1.2 For Saliva Step 1: Sample Collection

1. Vigorous mouthwash with 5 mL of sucrose solution (3%) for
1 min (vigorous movements of the tongue against the teeth
and all mucosa). Transfer to a disposable conical polypropylene
tubes (15 mL).
2. Add 3 mL of the TNE solution with ethanol and invert a few
times. The sample can be stored at 4 C for 1 week. For best
results, extract DNA as soon as possible.
Step 2: DNA Extraction

Day 1: Beginning of extraction
1. Complete the conical tube with autoclaved ultrapure water
(sufficient quantity to complete 15 mL).
2. Centrifuge for 10 min at 2000 g. The pellet is visible after the
centrifugation.
3. Discard the supernatant and add 5 mL of ethanol-free TNE
and vortex for 5 s.
4. Centrifuge for 5 min at 2000 g.
5. Discard the supernatant and add 1 mL of lysis solution, mixing
with the formed pellet.
6. Place the contents of the sample in a 2 mL tube and leave in the
heater dry block or incubator for 1 h at 50 C.
7. Add 10 μL of proteinase K (20 mg/mL) to the mixture (keep
proteinase on ice and spin for a few seconds before use).
8. Incubate overnight at 50 C.
Day 2
1. Turn refrigerated centrifuge on to 4 C.
2. Add 470 μL of the EDTA and ammonium acetate solution
(reverse the tube, at least, 20 times).
3. Centrifuge at 12,000 g for 10 min at 4 C.
4. Transfer 800 μL of the supernatant to a new two microtubes
(2 mL) previously identified.
5. Add 540 μL of isopropanol in each tube, and invert 20 times.
6. Centrifuge at 12,000 g for 5 min at 4 C (handle out of the
microtube).
7. Discard the supernatant and dry the border quickly on towel
paper.
8. Add 500 μL of 70% ethanol (cold) to the microtubes; resus-
pend until the pellet is released from the bottom of the tube.
Join the pair of microtubes in one single tube.
9. Centrifuge at 12,000 g for 5 min at 4 C.
10. Discard the ethanol. Dry the edges of the microtube with paper
towels.
11. Dry for 20 min (with microtube down) and 20 min with
microtube up (may be longer). The smell of ethanol must
be gone.
12. Resuspend the DNA in 50 μL TE buffer 1 (elution buffer)
and rest for, at least, 30 min. Observe dilution of the pellet.
13. Proceed to downstream application, or store the DNA at
20 C.
4.1.3 Determining Yield Quantify in a spectrophotometer (NanoVue™, Biochrom, or

and Quality of DNA NanoDrop™, Thermo Scientific). It permits rapid quantification
of undiluted DNA samples using a very low sample volume (1 μL),
checks purity and integrity, and also can be used to detect protein or
solvent contamination. To ensure good DNA quality, templates
should be analyzed by:
(a) Agarose gel electrophoresis using a known mass standard
where a visible band should be present on the gel at the
expected quantitated level
(b) Spectrophotometer to ensure:
OD260/OD280 range is between 1.8 and 2.0.
OD260/OD280 <1.8 may indicate protein contamination.
OD260/OD280 >2.0 may indicate RNA contamination.
4.2 DNA (a) Select the genomic region (access: Ensembl genome
Amplification [ensemblgenomes.org/] or UCSC genome [https://
genome.ucsc.edu/]).
4.2.1 Designing PCR
Primers (b) Choose the genomic region you want to amplify, and access
the Primer3 software for primer design [51, 52].
(c) Paste the genomic sequence, indicate which region to be
amplified with brackets, and pick primers.
i.e.: Exon 14—gene COL1A1
ttcttctgattcagGGTGAAGCTGGTCCCCAAGGGCCCCGA
GGCTCTGAAGGTCCCCAGGGTGTGCGTGGTGAG
CCTGGCCCCCCTGGCCCTGCTGGTGCTGCTGGC
CCTGCTgtaagtgtccccgactcagtgtcccctttgccactttctaacctcaga
gtccttgcctgttg
(d) Test with the basic local alignment search tool (BLAST) to
check if the primer is in the correct gene. The BLAST, http://
blast.ncbi.nlm.nih.gov/Blast.cgi, is a program that can detect
sequence similarity between a query sequence and sequences
within a database.
(e) Choose a primer annealing temperature. The melting temper-
ature (tm) of the primers determines the annealing tempera-
ture. The usual place to set the annealing temperature is about
2 C lower than the lowest Tm of the primers. Thus, if the
melting temperature is 59.1 C, for the forward primer, and
59.4 C for reverse primer, the annealing temperature should
be set at 57 C as a starting point. This temperature can be
changed up or down depending on the results.
4.2.2 Setting Up the PCR It is advisable to have pre-PCR and post-PCR rooms to avoid
Reaction contamination.
Prepare the appropriate number of reactions, plus 10% overage
(always change the tips, and care with contamination). The PCR
Table 3
PCR reactions setup
Component Volume (25 μL/reaction) Volume (50 μL/reaction) Final concentration

10 buffer 2.5 μL 5 μL 1
2 mM dNTPs 2.5 μL 5 μL 200 μM
10 μM forward primer 1 μL 2 μL 0.4 μM (0.05–1 μM)
10 μM reverse primer 1 μL 2 μL 0.4 μM (0.05–1 μM)
Taq DNA polymerase 0.2 μL 0.4 μL 2 units
a
PCR cosolvents Variable Variable Variable
DNA template Variable Variable <1000 ng
Nuclease-free water to 25 μL to 50 μL –
a
Optional
reaction can be performed for several final volumes. Below is a table

with setup PCR reactions for a final volume of 25 and 50 μL
(Table 3; insert here).
(a) The choice of the DNA polymerase is determined by the aims
of the experiment. There are varieties of commercially avail-
able enzymes to choose. They differ in their thermal stability,
processing, and fidelity. Some of DNA polymerase are
depicted in Table 4.
(b) Variety of PCR cosolvents have been utilized to increase the
yield, efficacy, and specificity of PCR amplifications [53].
Although these cosolvents are advantageous in some amplifi-
cations, it is impossible to predict which additive will be useful
for each primer. Some of the cosolvents currently in use are
listed in Table 5, along with the recommended testing ranges
(insert Tables 4 and 5 here).
(a) Mix the components thoroughly, and then centrifuge
briefly to spin down the contents and eliminate any air
bubbles.
(b) Transfer the appropriate volume of each reaction to pre-
vious identified microtubes of 0.2 mL, and add the DNA.
(c) Set the thermal cycling conditions. The number of PCR
amplification cycles should be optimized with respect to
the starting concentration of the target DNA.
Table 4
Commercially available enzymes
Blunt or
Proofreading Hot Product GC 50 !30 30 A
Enzyme 50 ! 30 exo Fidelity start size (kb) rich exo ends Primary application
Taq DNA <3 + 30 A Routine PCR—
polymerase primer extension
AmpliTaq +++ <5 + 30 A Routine PCR
0
AmpliTaq +++ + <5 + 3 A Hot start PCR
Gold
Vent DNA + ++ <6 + Blunt Routine high-fidelity
polymerase PCR
Platinum Tfi + + <2 + + 30 A Routine PCR and
DNA Y/N sequence
polymerase detection
Platinum Pfx + ++++ + <12 ++ Blunt High-fidelity PCR
DNA for cloning and
polymerase mutagenesis
Accuprime + +++ + <20 + Mixed Robust and long
Taq high PCR
fidelity
AmpliTaq ++ <2 + 30 A Multiplexing—high
Stoffel GC targets
fragment
DyNazyme + + <40 +++ + Mixed PCR difficult or
EXT DNA long—cloning—
polymerase microarrays
A standard cycle is recommended as follows, which can be

changed as required, increasing or decreasing the time and
temperature:

Initial denaturation ¼ 94 C5 min
8
>
> 94 C30 s
<
30 cycles Annealing temperature ðvariableÞ30 s
>
>
:
Extension temperature 72 C30 s ðsee Note 13Þ

Final extension ¼ 72 C5 min
Table 5
Cosolvents currently in use
Cosolvent Ranges Comments

Betaine Final Reduces the formation of secondary structure caused by
concentration: GC-rich regions
1.0–1.7 M
Bovine serum albumin 10–100 μg/mL A nonspecific enzyme stabilizer (BSA) which also binds
certain DNA inhibitors
7-Deaza- Ratio 3:1 Facilitates amplification of templates with stable
20 -deoxyguanosine dC7GTP: secondary structures when used in place of dGTP
(dC7GTP) dGTP
DMSO 2–10% 10% reduces Taq activity by 50%. Thought to reduce
secondary structure. Useful for GC-rich templates.
Presumed to lower the Tm of the target nucleic acids
Formamide 1–5% Improve the specificity of PCR at lower denaturation
temperatures
Glycerol 1–10% Improves the thermal stability of DNA polymerases.
Improves the amplification of high GC templates
Nonionic detergents: 0.1–1% Stabilizes Taq DNA polymerase. May suppress the
Triton X-100, Tween formulation of NP40 secondary structure. May
20 increase yield but may also increase nonspecific
amplification
T4 gene 32 protein 20–150 μg/mL Enhance PCR product yield and relieve inhibition
Tetramethylammonium Final To eliminate nonspecific priming. Also used to reduce
chloride (TMAC) concentration: potential DNARNA mismatch. Improves the
15–100 mM stringency of hybridization reactions
TMA oxalate 2 mM Decreases the formation of nonspecific DNA fragments
and increases PCR product yield
4.3 PCR Products This method should be only used if the band of interest is a single
Purification amplification product. All centrifugation steps are carried out at
17,900 g in a conventional tabletop microcentrifuge at room
4.3.1 Purification by
temperature.
Column (QIAquick® PCR
Purification, Qiagen) 1. Add 5 volumes buffer PB to 1 volume of the PCR reaction and
mix. If the color of the mixture is orange or violet, add 10 μL
3 M sodium acetate and pH 5.0, and mix. The color of the
mixture will turn yellow.
2. Place a QIAquick column in a provided 2 mL collection tube.
3. To bind DNA, apply the sample to the QIAquick column and
centrifuge for 30–60 s.
4. Discard flow-through and place the QIAquick column back in
the same tube.
5. To wash, add 0.75 mL buffer PE to the QIAquick column;

centrifuge for 30–60 s. Discard flow-through and place the
QIAquick column back in the same tube.
6. Centrifuge the QIAquick column once more in the provided
2 mL collection tube for 1 min to remove residual wash buffer.
7. Place each QIAquick column in a clean 1.5 mL
microcentrifuge tube.
8. To elute DNA, add 50 μL buffer EB (10 mM Tris–HCl, pH
8.5) or water (pH 7.0–8.5) to the center of the QIAquick
membrane and centrifuge the column for 1 min. For increased
DNA concentration, add 30 μL elution buffer to the center of
the QIAquick membrane, let the column stand for 1 min, and
then centrifuge.
9. If the purified DNA is to be analyzed on a gel, add 1 volume of
loading dye to 5 volumes of purified DNA. Mix the solution by
pipetting up and down before loading the gel.
4.3.2 Purification by When artifacts or more than one amplification product exist, isola-
Agarose PCR Gel Product tion of the band of interest is required by agarose gel electrophore-
sis. In addition, this technique allows the removal of primers and
dNTPs. The PCR product should be on a low-melting agarose gel,
cutting the corresponding band (try not to contaminate the band
with oligos and cut the smallest amount of agarose possible) and
melting the agarose. Finally, extract and purify the DNA with
successive steps of phenol/chloroform/isoamyl alcohol (25:24:1)
and precipitation with ethanol and salts. Alternatively, commercial
kits may be used for purification.
4.3.3 Purification by Treatment with exonuclease I degrades residual PCR primers while
ExoSAP-IT® PCR Protocol shrimp alkaline phosphatase (SAP) dephosphorylates the remaining
(Affymetrix) dNTPs. After inactivation of both enzymes, the PCR product can
be used for automatic sequencing and recommended for small PCR
products. It should not be applied in the case of PCR products
contaminated with secondary products.
1. Remove ExoSAP-IT reagent from 20 C freezer, and keep on
ice throughout this procedure (store ExoSAP-IT reagent in a
non-frost-free freezer).
2. Mix 5 μL of a post-PCR reaction product with 2 μL of
ExoSAP-IT reagent for a combined 7 μL reaction volume (see
Note 14).
3. Incubate at 37 C for 15 min to degrade remaining primers and
nucleotides.
4. Incubate at 80 C for 15 min to inactivate ExoSAP-IT reagent.
5. The PCR product is now ready for use in DNA sequencing,
SNP analyses, or other primer-extension applications.
6. Treated PCR products may be stored at 20 C until required.
4.4 Multiplex MLPA is the gold standard for DNA copy number quantification,
Ligation Probe such as deletions and duplications. MLPA can also be applied to
Amplification (MLPA) investigate the methylation status of DNA sequences. After the
MLPA reaction, fragment separation by capillary electrophoresis is
required, and the methods depend on the equipment used and
manufacturer’s instructions.
4.4.1 Materials 1. SALSA MLPA Probemix and corresponding product

description.
2. Ultrapure water.
3. TE (10 mM Tris–HCl pH 8.2 + 0.1 mM EDTA).
4. Thermocycler with heated lid (99–105 C).
5. PCR tubes, strips, or plates.
6. High quality formamide (e.g., Hi-Di Formamide, Applied
Biosystems).
7. Capillary electrophoresis equipment (Beckman Coulter or
Applied Biosystems).
8. Labeled size standard.
9. Polymers.
4.4.2 Methods DNA denaturation (day 1)

1. Use 0.2 mL microtubes.
2. Add 5 μL of sample DNA (50–250 ng) into each tube.
3. Place the microtubes in the thermocycler and select the MLPA-
DEN program (Table 6).
Hybridization reaction (day 1)
1. Vortex the buffer and probe before using.
2. Prepare the hybridization mix containing (by reaction):
(a) 1.5 μL of MLPA buffer.
(b) 1.5 μL of probe.
Table 6
MLPA program
MLPADEN MLPA LIG MLPA PCR (35 cycles)

98 C for 10 min; 25 C pause (pipetting) 54 C pause (pipetting) 95 C for 30 s
95 C for 1 min 54 C for 15 min 60 C for 30 s
60 C for 16–20 h 98 C for 5 min 72 C for 30 s
20 C infinite 72 C for 20 min
3. Mix well by pipetting or vortexing.

4. After denaturation, add 3 μL of the hybridization mix in
each tube.
5. Continue the program in the thermocycler: 1 min at 95 C and
16–20 h at 60 C (see Note 15).
Binding reaction (day 2)
1. Vortex the two ligase buffers before using them.
2. Prepare the binding mix containing (by reaction):
(a) 25 μL of water.
(b) 3 μL of Ligase buffer A.
(c) 3 μL Ligase buffer B.
(d) 1 μL Ligase-65 enzyme.
3. Mix by pipetting.
4. Start the MLPALIG program. When it reaches 54 C, pause,
and pipette 32 μL of mix in each tube (Table 6).
5. Continue the program (see Note 16).
PCR reaction (day 2)
1. Vortex the primer before using it. Heat the polymerase for 10 s
in hands to reduce viscosity.
2. Prepare the PCR mix containing (by reaction):
(a) 7.5 μL of water.
(b) 2 μL of SALSA PCR primer mix.
(c) 0.5 μL of SALSA polymerase.
3. Mix well by pipetting. Never vortex. Keep the mix on ice until
you use it.
4. At room temperature, add 10 μL of the PCR mix into each
tube. Mix by pipetting.
5. Start the MLPAPCR program (Table 6; insert here) (see
Note 17).
4.5 Methodologies Epigenetic gene regulation is highly dynamic and drives cell phe-
Used to Evaluate notype changes. Epigenetics refers to a change in a gene that is
Epigenetic passed on through cell division, but does not involve DNA
Modifications sequence change [54], instead, acting via chemical modifications
of the DNA molecule and histone proteins, i.e., DNA methyla-
tion/hydroxymethylation and histone modifications (acetylation,
methylation phosphorylation, and others). Chromatin accessibility
at gene regulatory elements, such as promoters, enhancers, and
locus control regions, is controlled by epigenetic modifications,
controlling the ability of transcription factors to bind to these
regions and regulate gene expression. Therefore, they are involved
in physiological and pathological conditions. Since epigenetic pro-

files are dynamic and potentially chemically modifiable, targeting
them brings an opportunity for novel therapies. There are several
methodologies and protocols for evaluating epigenetic changes;
some of them will be briefly described as whole epigenome
approach and gene-specific approach.
4.5.1 Whole Epigenome In addition to 5-methylcytosine (5mC), several different types of

Approach DNA base modifications in mammalian DNA have been discov-
ered, such as 5-hydroxymethylcytosine (5hmC), 5-formylcytosine
DNA (Hydroxy)methylation (5fC), and 5-carboxylcytosine (5caC). Their role on being func-
Profile (Array/Bead Chips) tional modifications or intermediates of active demethylation still
needs to be addressed. However, since bisulfite does not distinguish
between 5mC and 5hmC, new approaches have been proposed,
such as Infinium MethylationEPIC BeadChip array (Illumina) cou-
pled with the TrueMethyl™ Kit (Cambridge Epigenetix-CEGX,
Cambridge, UK). The traditional bisulfite conversion leads to
potential overestimation of 5mC levels due to the presence of
5hmC, the oxBS method generates a more accurate 5mC profile
by removing the confounding factor of 5hmC, and both chemical
modifications can be differentially identified by comparing methy-
lated CpG sites between the BS- and oxBS-treated samples [55].
The workflow comprises three main steps:
1. Oxidative bisulfite (oxBS) treatment: The same sample (1 μg) is
splitted in two equal reactions (500 ng each), one of which
undergoes chemical oxidation followed by bisulfite conversion
and the other undergoes mock oxidation (oxidant replaced by
water) followed by bisulfite conversion [56]. The oxidant solu-
tion converts 5hmC bases to its formyl derivative 5fC, which is
deaminated to uracil by bisulfite treatment; therefore, by selec-
tively oxidizing 5hmC to 5fC prior to bisulfite treatment, only
5mC remains unconverted by bisulfite treatment [57].
2. Bisulfite conversion: In the presence of bisulfite, 5fC is defor-
mylated and deaminated to uracil, after the following steps:
bisulfite-mediated conversion of unmethylated cytosines and
formylated cytosines to uracil, cleanup of bisulfite converted
DNA with magnetic beads, and desulfonation and elution of
DNA. The final volume of the recovered TrueMethyl® con-
verted DNA sample is 15 μL, and the expected mass depends
on the initial input. The converted DNA sample is compatible
with downstream next-generation sequencing or arrays/bead
chips analysis.
3. (Hydroxy)methylation array: The MethylationEPIC BeadChip
(Illumina) follows a workflow that does not require PCR. Its
low sample input requirement (as low as 250 ng) enables
analysis from limited DNA sources. The results can be obtained
after the BeadChips are scanned on iScan (Illumina). The

scanner records high-resolution images of the light emitted
from the fluorophores.
Data Processing (Hydroxy)methylation levels can be calculated using associated

Genome Studio software or performed using the R statistical lan-
guage version 3.3.0 and various packages of the Bioconductor
project [56]. OxBS-generated methylation profiles must be sub-
tracted from standard BS-only methylation profiles for the detec-
tion of hydroxymethylated cytosine positions within the genome
[55–57].
The Infinium Methylation EPIC BeadChip enables to study
DNA (hydroxy)methylation quantitatively at 830,000 sites, includ-
ing CpG sites, RefSeq genes, ENCODE open chromatin,
ENCODE transcription factor binding sites, and FANTOM5
enhancers. Since enhancers possess the highest enrichment of
5hmC among all genomic elements [58], a substantial proportion
of potentially 5hmC-dependent functional genomic elements are
detected by the Infinium Methylation EPIC array.
Materials 1. Pipettes and pipette tips with aerosol barriers.

2. 0.2, 1.5 and 2.0 mL polypropylene tubes.
3. Thermal cycler with heated lid.
4. Variable speed microcentrifuge.
5. Tube adapters for 0.2 mL thermal cycling tubes.
6. Heat block, thermomixer, or heated orbital incubator able to
maintain temperatures at 37 and 60 C and shake up to 1400
RPM (e.g., Eppendorf® Thermomixer Comfort).
7. Ice bucket.
8. Plastic microcentrifuge tube holder or rack.
9. Magnetic separation rack (e.g., Invitrogen™ Dynamag-2-
magnet P/N: 12321D).
10. HPLC grade acetonitrile.
All steps can be found in the TrueMethyl™Kit User Guide and
in the Infinium HD Assay Methylation Protocol Guide.
Methylated and MeDIP and hMeDIP-seq are sequencing-based methods classified

Hydroxymethylated DNA as enrichment-based method, since it does not involve bisulfite
Immunoprecipitation conversion and is able to detect methylated cytosines. Taiwo
(MeDIP and hMeDIP)-seq et al., 2012 proposed an optimized MeDIP-seq protocol that
requires 50 ng of genomic input DNA. The main difference is
that the generic library preparation is split into two phases, with
the MeDIP procedure sandwiched between two library preparation
steps. The low-input DNA makes MeDIP-seq suitable for studies
involving minute clinical samples, microdissected tissues, and rare

cell types. Aiming targeting the vast majority of the methylome, the
MeDIP-seq protocol uses antibodies directed against mC or mCG
to precipitate methylated DNA fragments, providing methylomes
at 100–300 bp resolution. After DNA fragments are precipitated,
the library preparation is carried out, and after adapter ligation, the
MeDIP is performed. After purification and quantification of the
MeDIP-seq library, a certain amount of MeDIP-seq library is
sequenced. Details of the MeDIP-seq protocol can be found in
Taiwo et al., 2012. The hMeDIP recognizes hmC, with the appro-
priate antibody [59] and possibly 5-formylcytosine and
5-carboxylcytosine as well [60]. hMeDIP-seq is carried out simi-
larly to MeDIP-seq protocol and has become an invaluable tool for
determining both locus-specific and genome-wide profiles of
5hmC in mammalian DNA [61]. One of their limitations is that
both MeDIP-seq and hMeDIP-seq have relatively low resolution
and cannot quantitatively determine 5mC and 5hmC abundance in
a single base resolution manner [58]. Skvortsova et al. (2017)
showed that for samples containing DNA with 5hmC in abundance
(>10%), hMeDIP-seq shows high specificity. On the other hand,
for very low levels of 5hmC (up to 4%), hMeDIP-seq signals cannot
be used as the representative of the actual 5hmC distribution
without further validation.
Materials 1. End repair, A-tail, and adapter ligation module (NEB, cat.
no. E6050L, E6053L and E6056L).
2. High-fidelity Phusion polymerase (5; NEB, cat.
no. M0530L).
3. dNTP mix (NEB, cat. no. N0447L).
4. Auto-MeDIP kit (Diagenode, cat. no. AF-Auto01-0016 or
AF-Auto01-0100). * MeDIP can also be performed manually
using the MagMeDIP kit (Diagenode, cat. no. mc-magme-
A10) or other kits/reagents. In this case, we advise taking
care with respect to accurate and consistent timing of incuba-
tion steps, as well as performing lengthy and thorough bead
washing steps.
5. IPure kit (Diagenode, cat. no. AL-Auto01-0100) *MeDIP can
also be performed manually using the MagMeDIP kit (Diag-
enode, cat. no. mc-magme-A10) or other kits/reagents. In this
case, we advise taking care with respect to accurate and consis-
tent timing of incubation steps, as well as performing lengthy
and thorough bead washing steps.
6. Sequencing adapter, primer oligos, and QC primers (Sigma).
7. MinElute Gel extraction kit (Qiagen, cat. no. 28004).
8. Lambda (λ)-DNA (NEB, cat. no. N3011S).
9. SssI CpG methyltransferase (NEB, cat. no. M0226S).
10. Agencourt Ampure XP (Beckman Coulter, cat. no. A63881).

11. Ethanol, isopropanol, and sodium acetate (Sigma-Aldrich, cat.
E7023, I9516 and S7899).
12. NuSieve low-melting point Agarose (Lonza, cat. no. 50084).
13. Ladder (50 bp; NEB, cat. no. N3236L or similar).
14. Tris Acetate EDTA (TAE, 50; Severn Biotech, cat.
no. 206001-10).
15. Ethidium bromide and β-mercaptoethanol (Sigma, cat.
no. E1510 and M7522).
16. Loading dye (Fermentas, cat. no. R0631).
17. AllPrep DNA/RNA Micro Kit (Qiagen, cat. no. 80284).
18. Agilent DNA and Agilent High-Sensitivity kit (Agilent, cat. no.
5067-1504 and 5067-4626).
19. SYBR qPCR master mix (2; Eurogentec, cat. no. RT-sy2x-
03-Woub or similar).
20. Ultrapure water and HCl.
21. DNA Clean and Concentrator-5 columns (Zymo).
22. RLT plus lysis buffer (Qiagen).
23. Microcentrifuge tubes (Eppendorf 0.5 mL, cat.
no. 0030108.035; 1.5 mL, 0030108.0510; 2 mL
0030108.078 or similar).
24. PCR tubes with flat cap (0.2 mL).
25. PCR tubes and PCR tube caps (8 strips and 12 strips).
26. IP-Star tips (Diagenode, cat. no. WC-001-1000).
27. qPCR plate (96 well) and plate seal.
28. Benchtop mini centrifuge with 0.2 mL tube adapter (Jencons
PLC, cat. no. C1301-230V).
29. Pipettes and pipette filter tips.
30. Vortex.
31. Transilluminator.
32. Microwave oven.
33. Magnetic separation rack.
34. Single-pouch Sterile 10A scalpel (Swann-Morton or
equivalent).
35. Thermocycler machine (96-well plate), real-time PCR, and gel
electrophoresis system.
36. Gel comb (3 mm; Owl Separation Systems or equivalent).
37. NanoDrop spectrophotometer.
38. Incubator oven (Techne or equivalent).
39. Bioruptor UCD-200 sonicator (Diagenode).
40. Refrigerated microcentrifuge.

41. Heat block.
42. Agilent Bioanalyzer (Agilent).
43. Illumina GAIIx (Illumina).
44. SX-8G IP-Star (Diagenode). The SX-8G IP-Star is used for
automated protocol; however, generic liquid handlers can be
programmed to carry out the MeDIP and hMeDIP-seq as
detailed in the MagMeDIP protocol (Diagenode).
Open Chromatin Regions: Each nucleus of each single cell of the human body contains almost
Assay for Transposase two meters of DNA. This packaging has inactive regions with
Accessible Chromatin repressing epigenetic marks and biologically active regions that
(ATAC) are accessible for transcription machinery when needed. Dynamic
epigenetic marks and mechanisms drive these changes in chromatin
architecture. Open chromatin regions by ATAC-seq followed by
next-generation sequencing is a method for mapping these open
(accessible) chromatin regions, identifying active regulatory ele-
ments and transcription factor binding sites. This method identifies
promoters, enhancers, and locus control regions and probes DNA
accessibility with hyperactive Tn5 transposase:
1. Tn5 transposase cuts and ligate adapters into accessible regions
of chromatin.
2. Immediately following transposition, samples are purified and
eluted.
3. The transposed DNA fragments are amplified aiming to gener-
ate libraries that are minimally PCR amplified. Purify amplified
library. The purified library is eluted in 20 μL elution buffer
(10 mM Tris buffer, pH 8).
4. The quantification of the libraries can be done using Kapa
Library Quantification Kit (Kapa Biosystems).
Sequencing reads can be used to infer regions of increased
accessibility, as well as to map regions of transcription factor bind-
ing and nucleosome position [62]. The different populations of
reads (shorter or consistent with the length of DNA protected by a
nucleosome) provide information about the positions of nucleo-
somes and nucleosome-free regions. After ATAC-seq, the genome-
wide data will inform on which regulatory elements are active in the
analyzed samples and in addition, which genes are potentially regu-
lated and which transcription factors are potentially involved under
a specific situation.
Materials 1. Phosphate-buffered saline (PBS) without calcium and

magnesium.
2. Molecular biology grade IGEPAL CA-630.
3. Lysis buffer (10 mM Tris–HCl, pH 7.4, 10 mM NaCl, 3 mM

MgCl2, 0.1%).
4. 2 TD (2 reaction buffer, Illumina cat. no. FC-121-1030).
5. TDE1 (Nextera Tn5 Transposase, Illumina cat.
no. FC-121-1030).
6. Qiagen MinElute PCR Purification Kit (Qiagen cat.
no. 28004).
7. NEBNext High-Fidelity 2x PCR Master Mix (New England
Labs cat. no. M0541).
8. 25 μM Custom Nextera PCR Primer 1.
9. 25 μM Custom Nextera PCR Primer 2.
10. 100 SYBR Green I (Invitrogen cat. no. S-7563).
11. 0.2 mL PCR tubes.
12. PCR thermal cycler.
4.5.2 Gene-Specific This approach might be employed for gene-specific studies and for
Approach: Quantitative PCR further validation of global approaches. The workflow is as follows:
Assay for Gene-Specific
1. T4-β-glucosyltransferase (T4-BGT) treatment: gDNAs are initi-
DNA Methylation and
ally treated with T4-BGT, adding glucose moiety to 5-hmC to
Hydroxymethylation
distinguish among DNA methylation and hydroxymethylation.
For each sample, three tubes containing 400 ng gDNA each are
treated with 1X NE buffer 4, 40 mM UDP glucose, and
T4-BGT (1 unit) and incubated at 37 C for 1 h, followed by
10 min at 65 C.
2. Restriction enzymes digestion: samples are digested with MspI or
HpaII restriction enzymes or H2O (control) at 37 C for 1 h.
Tubes containing the HpaII restriction enzyme are submitted
to additional incubation for 10 min at 65 C, for enzyme
inactivation.
3. Amplification of the target region: DNA is subjected to amplifi-
cation cycles using primers at proper concentration and anneal-
ing temperature for each target region. Analyses are conducted
in the PCR apparatus using a SYBR Green dye. The copy
number might be determined by using a standard curve, and
samples can be normalized by dividing the copy number of
reactions by the copy number of the control reaction. The
comparative Ct method might be also employed, and samples
are normalized by setting the control reaction as calibrator.
MspI and HpaII restriction enzymes recognize the same
sequence (CCGG); however, HpaII cuts the unmodified site,
and any epigenetic modification (5mC, 5hmC, 5ghmC) blocks
cleavage. MspI recognizes and cuts 5mC and 5hmC, but does
not cut the 5ghmC modification. Primers should be designed
on epigenetic regulatory regions such as DNaseI
hypersensitivity cluster sites, layered by histone modifications

marks, CpG regions, and transcription factor binding sites that
contain the CCGG sequence.
Materials 1. T4-BGT, 1X NE buffer 4, and UDP glucose (New England

Biolabs cat. no. M0357S).
2. MspI (New England Biolabs cat. no. R0106S).
3. HpaII (New England Biolabs cat. no. R0171S).
4. FastStart Essential DNA Green Master or similar (Roche cat.
no. 6402712001).
5. 96 multi-well plates and plate seal.
6. Real-time PCR system thermal cycler.
5 Notes
1. DNA salting out method. This method seems to be more

efficient and with less generation of toxic waste.
2. Reagents. Prepare all solutions using ultrapure water and ana-
lytical grade reagents. Prepare and store all reagents at room
temperature (unless indicated otherwise). All solutions must be
autoclaved.
3. RBC lysis solution. To prepare 100 mL of solution, add 0.5 mL
of MgCl2 1 M and 0.2 mL of EDTA 0.5 M and complete with
ultrapure water.
4. Cell lysis solution for blood sample. To prepare 100 mL of
solution, add 1 ml of Tris 1 M, 0.2 mL of EDTA 0.5 M, and
10 mL of SDS 10% and complete with ultrapure water.
5. Protein precipitation solution. Add 57.81 g of ammonium ace-
tate and complete to 100 mL of ultrapure water.
6. TE buffer 1. To prepare 10 mL, add 295 μL of Tris–HCl
0.34 M (pH 7.4) and 72 μL of EDTA 0.14 M (pH 8.0) and
complete with ultrapure water. Mix briefly.
7. TNE solution. To prepare the solution, you have to add in a
bottle: 1.5 mL of Tris 0.34 M, 1.5 mL of NaCl 1 M, 1.5 mL of
EDTA 0,14 M, and 25.5 mL of ultrapure water. Mix briefly.
8. TNE solution with ethanol. Add 1.5 mL of Tris 0.34 M, 1.5 mL
of NaCl 1 M, 1.5 mL of EDTA 0.14 M, 5.7 mL of ultrapure
water, and 19.8 mL of ethanol. Mix briefly.
9. Lysis solution for saliva samples. Add 2 mL of SDS 5%, 588 μL of
Tris 0.34 M, 715 μL of EDTA 0.14 M, and 16.7 mL of
ultrapure water.
10. Solution of EDTA and ammonium acetate. Add 70 μL de
EDTA 0.14 M and 10 mL of ammonium acetate.
11. Most suppliers of Taq polymerase provide buffer solution opti-

mized for their own particular enzyme.
12. Appropriate PCR product size. Small fragments can behave as
primers and are difficult to purify with good performance. A
product size of at least 150 bp is recommended.
13. Polymerase extension step. The amplicon length determines the
cycle step duration. Thirty seconds for every 500 bp of product
is recommended. Thus, for a predicted 2300 bp amplicon size,
an extension step with 2.5 min duration should be necessary.
14. When treating PCR product volumes greater than 5 μL, simply
increase the amount of ExoSAP-IT reagent proportionally.
15. After the reaction do not open the tubes in the thermocycler
room. To avoid contamination, use different micropipettes to
make the MLPA and for the manipulation of the PCR
products.
16. The product of the binding reaction can be stored at room
temperature for a few hours, or 4 C for up to a week.
17. The PCR reaction product can be stored at 4 C for up to
1 week. For longer periods, store between 25 and 15 C.
The PCR product should be stored in a dark box or wrapped
in foil.
6 Part 2
6.1 Establishment Patients with genetic dental anomalies frequently require early
and Characterization treatments, which can involve dental extraction and periodontal
of Human Cell Culture surgeries. These procedures provide tissues that may be useful for
for Functional Studies primary cell culture establishment. As primary cells retain many of
native cellular functions in vitro, primary culture of patient with
dental anomalies represent a convenient tool for studying the
behavior of cells carrying a pathogenic variant when compared
with primary cells from healthy controls [63, 64]. Comparison
between mutated cells and normal cells contribute to better under-
standing the impact of a gene pathogenic variant and can bring
benefits for overall treatment and rehabilitation. Thus, this section
will describe methods of isolation and characterization of primary
culture of gingival, pulp, and periodontal ligament tissues from
patients with genetic dental anomalies.
7 Materials
7.1 Cell Isolation and Scalpel with a #15 blade.

Culture Establishment Sterilized cover slips.
7.1.1 Instruments and Periodontal curette.
Materials Tweezers.
Tooth extraction forceps.
10 mL syringe.
Chisel.
Mallet.
Ice bucket.
70 μm cell strainer (SPL Life Science).
Polypropylene conical tube (50 mL).
35, 60, and 100 mm culture dish.
Hemocytometer.
7.1.2 Reagents Phosphate-buffered saline (PBS) pH 7.4 without calcium and

magnesium.
Rinse solution: Phosphate-buffered saline (PBS) with antibioti-
c–antimycotic 100, liquid (Gibco™, ThermoFisher) (see
Note 1).
Digestion enzyme solution: DMEM with 3 mg/mL collagenase
type I (Sigma-Aldrich) and 4 mg/mL dispase II (Sigma-
Aldrich). Sterilize the solution by filtration through a 0.2 μm
membrane. Prepare just before use.
Complete growth medium: Dulbelcco’s Modified Eagle Medium
(DMEM) with 50 U/mL of penicillin, 50 μg/mL streptomy-
cin, and 20% fetal bovine serum (FBS).
Complete culture medium: Dubelcco’s Modified Eagle Medium
(DMEM) with 50 U/mL of penicillin, 50 μg/mL streptomy-
cin, and 10% fetal bovine serum (FBS).
Trypsin/EDTA (0.05% trypsin and 0.02% EDTA in PBS without
calcium and magnesium, pH 7.4). Sterilize the solution by
filtration through a 0.2 μm membrane.
Freezing solution: FBS with 10% DMSO.
7.2 Cell Cover slips.

Characterization Phosphate-buffered saline (PBS) pH 7.4 without calcium and
7.2.1 Actin Cytoskeleton magnesium.
Stained with Phalloidin 4% paraformaldehyde in PBS.
0.5% Triton X-100 (Sigma-Aldrich) in PBS.
Fluoroshield mounting medium with DAPI (Southern Biotech).
Fluorescein phalloidin conjugate (Invitrogen) prepared in PBS
(1:200 dilution). Prepare just before use.
Glass slides.
Fluorescent microscope.
7.2.2 Metabolic Activity 96 multi-well plates.

Assessed by 5 mg/mL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo-
3-(4,5-Dimethylthiazol-2- lium bromide (MTT) (Sigma-Aldrich) in PBS. Prepare just
yl)-2,5 Diphenyltetrazolium before use.
Bromide (MTT)
Dimethyl sulfoxide (DMSO).
Plate reader spectrophotometer.
7.2.3 Migration Capacity 10 μg/mL fibronectin (Sigma-Aldrich) in PBS. Prepare just

Assessed by Scratch before use.
Wound Healing Assay 200 μL pipette tip.
Phosphate-buffered saline (PBS) pH 7.4.
Inverted microscope with camera.
8 Methods
8.1 Cell Isolation and Routine techniques for working under sterile conditions (use of
Culture Establishment sterile media and instruments and working in a flow cabinet) should
be used.
The establishment of primary culture from human gingival,
pulp, and periodontal ligament tissues is described below. All tissue
collection should be made after informed consent approval.
8.1.1 Primary Explant 1. Collect the extracted tooth and store in a sterile conical tube
Culture of Gingival with 20 mL of complete growth medium in an ice bucket.
Fibroblasts (Fig. 1) Transport to the laboratory with minimal delay.
2. Discard the transport medium; transfer the sample to a sterile
100 mm Petri dish and rinse twice the tissue with rinse solution
to remove blood and debris.
3. Fragment the tissue into 3–5 mm pieces with a sterile scalpel
blade.
4. Transfer the fragments to sterile 35 mm Petri dish and cover

the sample with sterilized cover slips (see Note 2).
5. Add 3 mL of complete growth medium to each dish and
culture at 37 C in humidified atmosphere with 5% CO2.
6. Culture the explants undisturbed for 5 days and, then, replace
the medium taking care not to dislodge the explants. Replace
the medium twice weekly thereafter. A volume of 2 mL per dish
is suitable.
7. Check cells outgrowth at 7–15 days. Culture until near conflu-
ence (90%), generally 4–6 weeks post plating.
8.1.2 Isolation and 1. Collect the extracted tooth and store in a conical tube with
Primary Explant Culture of 20 mL of complete growth medium in an ice bucket. Transport
Pulp Fibroblasts (Figs. 2 to the laboratory with minimal delay.
and 3) 2. Discard the transport medium; transfer the sample to a sterile
100 mm Petri dish and rinse twice the tooth with rinse solution
to remove blood and debris.
3. Use a sterile blade to remove gingival and periodontal ligament
tissue and discard.
4. Make a longitudinal furrow just below the enamel–cementum
junction with a sterilized diamond drill, under continuous
irrigation, without reaching the pulp tissue. Then, fracture
the teeth (see Note 3) (Fig. 2).
5. Pull out the pulp using tweezers and dentinal excavator, trans-
fer to a 100 mm Petri dish, and rinse with 5 mL of rinse
solution and then with 5 mL of growth medium.
6. Fragment the tissue into 3–5 mm pieces with a sterile scalpel
blade.
7. Transfer the fragments to sterile 35 mm Petri dish and cover
the sample with sterilized cover slips (see Note 2).
8. Add 3 mL of complete growth medium to each dish and
culture at 37 C in humidified atmosphere with 5% CO2.
9. Culture the explants undisturbed for 5 days and replace the
medium, taking care not to dislodge the explants. Replace the
medium twice a week. A volume of 2 mL per dish is convenient.
10. Check for outgrowth of cells at 7–15 days. Culture until near
confluence (90%), generally 4–6 weeks post plating.
8.1.3 Isolation and 1. Collect the extracted tooth and store in a conical tube with
Primary Digestion Culture 20 mL of complete growth medium in an ice bucket. Transport
of Periodontal Ligament to the laboratory with minimal delay.
Fibroblasts (Fig. 4) 2. Discard the transport medium, transfer the tooth to a sterile
100 mm Petri dish, and rinse twice the material with rinse
solution to remove blood and debris.
Fig. 2 Steps to isolate gingival fibroblasts. (a) Gingival tissue rinsed with rinse solution. (b, c) Tissue
fragmentation using surgical blade. (d, e) Fragments are transferred to 35 mm culture dish and stabilized
with cover slips. (f) Addition of growth culture medium
3. Collect periodontal ligament tissue from the middle third of

the root using a surgical blade or periodontal curette. A twee-
zer can be used to stabilize the tooth. Transfer the sample to
sterile 60 mm Petri dish (see Note 4).
4. Add 3 mL of digestion enzyme solution and incubate at 37 C
for 1 h.
5. Collect the solution with a 10 mL syringe with a hypodermic
needle, and release an aspirate the solution for several times to
dissociate the cells.
Fig. 3 Isolate pulp tissue from unerupted third molar. (a) Panoramic radiograph showing unerupted third molar.
(b) Photograph tooth after removing gingival and periodontal ligament debris. (c) A longitudinal furrow just
below the enamel–cementum junction. (d) Photograph of pulp tissue
6. Filter the cell suspension with a 70 μm cell strainer, centrifuge

at 750 g for 5 min, and then discard the supernatant.
7. Resuspend the pellet with 5 mL of complete growth medium
and transfer to a 60 mm Petri dish culture at 37 C in humidi-
fied atmosphere with 5% CO2. Replace the medium after
5 days. Then, replace it twice weekly until attaining confluence,
generally 4–6 weeks post establishment.
8.2 Cell Culture This section describes the methods for expanding primary culture
Expansion and and cryopreservation.
Cryopreservation
1. Remove the medium.
2. Gently rinse the cells one time with 1 mL of PBS.
3. Add 1 mL of trypsin–EDTA solution to each dish and incubate
for until 5 min at 37 C.
4. Remove the dishes from the incubator and examine under the
microscope. The cell must become detached from the plate
Fig. 4 Steps to isolate pulp fibroblasts. Pulp tissue rinsed with (a) rinse solution and (b) with growth medium.
(c) Tissue fragmentation using surgical blade. (d) Fragment fixation with cover slips. (e, f) Addition of growth
culture medium
surface, with rounded, highly refractile cell bodies floating in

the trypsin–EDTA solution.
5. Add 1 mL of complete culture medium and transfer the cells to
50 mL conical tubes.
6. Centrifuge at 250 g for 5 min at room temperature to pellet

the cells.
7. Discard the supernatant, aspirating the medium or inverting
the tubes.
8. Add freshly regular medium and transfer the cells for 100 mm
dishes. For each 35 mm dish, transfer to one 100 mm dish.
9. Expand the cells until near confluence (around 90–95%), repla-
cing the medium twice a week, and then, harvest the cells as
described in this subheading, steps 1–4, using 5 mL of PBS to
rinse the cells.
10. When most of the cells have become detached from the plate
surface, add 4 mL of complete growth medium, mix the cells,
and transfer 10 μL of the mixed cell suspension to a hemocy-
tometer for cell counting.
11. Centrifuge at 250 g for 5 min at room temperature to pellet
the cells.
12. Discard the supernatant, aspirating the medium or inverting
the tubes.
13. Resuspend the cell pellet in freezing solution. Usually a density
of 2 106 cells per ml is suitable.
14. Transfer 1 mL of cell suspension to a cryoampoule.
15. Close ampoules tightly and freeze at 1 C/min to 80 C in a
cell freezing container (Mr Frosty, Nalgene).
16. After 24 h, transfer the cells to liquid nitrogen for long-term
storage.
8.2.1 Cell This section describes three methods to aid in the characterization
Characterization of mutated primary cultures when compared with those from
healthy patients (see Note 5). The use of cell subcultures up to six
passages is recommended. Morphology observation should be per-
formed by using phase-contrast inverted microscope (Fig. 5a).
Phalloidin staining is a useful tool to visualize actin filaments,
which is involved with diverse cellular functions, including adhe-
sion and contraction, cell–cell contact regulation, polarity, and cell
shape control. MTT activity assay is a colorimetric assay which
assesses mitochondrial activity and cell viability [65]. Wound heal-
ing assay is performed to evaluate migration capacity of the cells, by
creating an artificial scratch [66].
8.2.2 Actin Cytoskeleton 1. Add a sterile glass cover slips inside the culture plate.
Stained with Phalloidin 2. Plate the cells on sterile cover slips at a 2.6 103 cell per cm2,
(Fig. 5b) to reach a 50% confluence after 1 or 2 days post plating.
3. When the cell reached 50% confluence, discard the medium
and rinse cells one time with PBS (see Note 6).
Fig. 5 Steps to isolate periodontal ligament fibroblasts. (a) Tooth rinsed with washing solution. (b, c) Collection
of periodontal ligament from middle third of the root with a periodontal curette. (d) Addition of digestion
enzyme solution. (e) Filtration with 70 μm cell strainer. (f) Addition of storage medium
4. Fix cells with 4% paraformaldehyde, during 10 min at room

temperature.
5. Rinse with PBS three times for each 5 min.
6. Transfer the cover slips to 35 mm sterile Petri dish.
7. Permeabilize cells with Triton X-100 0.5% for 10 min at room
temperature.
8. Rinse with PBS three times for each 5 min.
9. Dilute 1:200 dilution in PBS of fluorescein phalloidin
(200 units/mL) and put on the slides, a minimum of 200 μL
per slide, during 50 min at 4 C in humified chamber, pro-
tected from light (see Note 7).
10. Rinse with PBS three times for each 5 min in light absence.
11. Air-dry the slides and mount on glass slides with fluoroshield
mounting medium with DAPI to preserve fluorescence and
counterstain DNA.
12. Visualize under fluorescent microscope and keep shielded from
light to prevent photobleaching. Store at 4 C up to 6 months.
8.2.3 Metabolic Activity 1. Plate the cells at a 1.56 104 cell per cm2 on 96-well plates (see
Assessed by MTT Note 8).
2. After each predetermined period, prepare MTT solution
(5 mg/mL) in PBS and add 10 μL per 90 μL of medium
(10 dilution) on each well.
3. Incubate the plates at 37 C and protect from light during 4 h.
4. Aspirate the medium and add 100 μL of DMSO to solubilize
the formazan crystals (make up and down with a pipette at least
ten times to complete dilute the crystals).
5. Read the absorbance on a microplate reader at a wavelength of
570 nm.
6. Collect the data, and construct a curve (absorbance time) to
compare the metabolic activity of primary culture from dental
anomaly patients and the control cultures.
8.2.4 Migration Capacity 1. Previously, dilute fibronectin at 10 μg/mL in PBS and coat the
Assessed by Scratch plate bottom surface with a minimal volume. Incubate plates
Wound Healing Assay overnight at 4 C in a closed sterile container.
(Fig. 7) 2. Aspirate the fibronectin solution and allow to air-dry for at least
45 min at room temperature. Proceed with the assay or store
the plates at the refrigerator (see Note 9).
3. Plate the cells at a 5.26 104 cell per cm2 (see Note 10).
4. After the cells attain confluence, scrape the cell monolayer in a
straight line with a p200 pipet tip to create a “scratch” (Fig. 6)
(see Note 11).
5. Rinse one time with PBS to remove the debris, and then replace
with medium.
6. Create markings to be used as reference points close to the
scratch to obtain the same field during the image acquisition
(see Note 12).
7. After the reference points are made, acquire the first image of
the scratch (time 0 h).
8. Place the plate in a tissue culture incubator at 37 C, until the
second photograph acquisition (see Note 13).
9. After the incubation, place the plate under a phase-contrast
microscope, match the reference point, align the photographed
Fig. 6 Cell morphology. (a) Spindle-shaped morphology of pulp fibroblasts

observed by phase-contrast inverted microscope. (b) Phalloidin staining.
Phalloidin binds F-actin, while FITC provides green fluorescence, allowing to
perform quantitative analysis. Nucleus is stained with DAPI
region, and acquire the subsequent images, until cell migration

to close the scratch.
10. The images acquired for each sample can be further analyzed
quantitatively by using computing software of choice. For each
image, distances between one side of scratch and the other can
be measured at certain intervals using a freeware (http://rsb.
info.nih.gov/ij/).
11. The rate of wound healing of each time compared with time
0 h can be determined by the following formula, with “A”
being the average of the distances in time 0 h and “B” the
average of the distances in the other hours:
[(A – B) 100%] A.
Fig. 7 Step for scratch wound healing assay. (a) Positioning the ruler as a guide. (b) Scrape the cell monolayer
in a straight line with a p200 pipet tip. (c) Remove the medium to rinse the cells with PBS to remove the debris
9 Notes
1. To prepare 100 mL of rinse solution, add 1 mL of antibioti-

c–antimycotic (Gibco™, ThermoFisher) in 100 mL of
sterile PSB.
2. Alternatively, a 6-well plate can be used.
3. Irrigation should be done in abundance, to reduce heat
buildup and to avoid damaging the pulp tissue. Alternatively,
to remove the dental pulp from teeth, stabilize the tooth using
a forceps and split the tooth using a chisel and mallet.
4. Preferably, use a curette without cutting. This minimizes the
amount of debris.
5. In order to compare primary culture from patient with dental
anomalies and health patient, establish culture of the two tis-
sues (normal mutated) and follow the experiments in
parallel.
6. In this point, it is important that mutated and normal cells have
similar density. Difference in the density between two culture
can result to I actin cytoskeletal differences.
7. Put the slides in a covered container to prevent evaporation and
light exposure.
8. Use one plate for each time that will be assessed. For example, if
the metabolic activity will be assessed on days 1, 3, 7, and
10, plate cells on four 96-well plates. Replace the medium
three times weekly. Use 100 μL of medium per each well.
9. Fibronectin-coated cultureware can be stored for 2–4 weeks at
2–8 C in a closed sterile container or sterile sealable bags.
10. Usually, with this density, 1–2 days may be needed for the
formation of cell monolayer.
11. It is important to create scratches of approximately similar size
in the assessed cells and control cells to minimize any possible
variation caused by the difference in the width of the scratches.
The use of a ruler makes it easy to get a stright line.
12. The reference points can be made by etching the dish lightly
with a razor blade on the outer bottom of the dish or with an
ultrafine tip marker. Alternatively, cover the microscope table
with adhesive tapes and create reference points on them.
13. For gingival, pulp, and periodontal ligament culture, a 12-h
time frame may be adequate. The dishes can be taken out of the
incubator to be examined periodically and then returned to
resume incubation.
References
1. Thesleff I (2003) Epithelial-mesenchymal sig- inherited anomalies affecting the dentition.
nalling regulating tooth morphogenesis. J Cell Wiley Interdiscip Rev Dev Biol 2(2):183–212
Sci 116(9):1647–1648 8. Vastardis H, Karimbux N, Guthua SW et al
2. Jussila M, Thesleff I (2012) Signaling networks (1996) A human MSX1 homeodomain mis-
regulating tooth organogenesis and regenera- sense mutation causes selective tooth agenesis.
tion, and the specification of dental mesenchy- Nat Genet 13(4):417–421
mal and epithelial cell lineages. Cold Spring 9. Prasad MK, Geoffroy V, Vicaire S et al (2016)
Harbor Perspect Biol 4(4):a008425 A targeted next-generation sequencing assay
3. Brook A (2009) Multilevel complex interac- for the molecular diagnosis of genetic disorders
tions between genetic, epigenetic and environ- with orodental involvement. J Med Genet 53
mental factors in the aetiology of anomalies of (2):98–110
dental development. Arch Oral Biol 54:S3–S17 10. Vastardis H (2000) The genetics of human
4. Townsend G, Bockmann M, Hughes T et al tooth agenesis: new discoveries for understand-
(2012) Genetic, environmental and epigenetic ing dental anomalies. Am J Orthod Dentofac
influences on variation in human tooth num- Orthop 117(6):650–656
ber, size and shape. Odontology 100(1):1–9 11. Pagnan NAB, Visinoni ÁF (2014) Update on
5. Wang J, Sun K, Shen Y et al (2016) DNA ectodermal dysplasias clinical classification. Am
methylation is critical for tooth agenesis: impli- J Med Genet A 164(10):2415–2423
cations for sporadic non-syndromic anodontia 12. Ye X, Attaie AB (2016) Genetic basis of non-
and hypodontia. Sci Rep 6:19162 syndromic and syndromic tooth agenesis. J
6. Bailleul-Forestier I, Berdal A, Vinckier F et al Pediatr Genet 5(4):198–208
(2008) The genetic basis of inherited anoma- 13. Sarkar T, Bansal R, Das P (2014) Whole
lies of the teeth. Part 2: syndromes with signif- genome sequencing reveals novel
icant dental involvement. Eur J Med Genet 51 non-synonymous mutation in ectodysplasin a
(5):383–408 (EDA) associated with non-syndromic
7. Cobourne MT, Sharpe PT (2013) Diseases of X-linked dominant congenital tooth agenesis.
the tooth: the genetic and molecular basis of PLoS One 9(9):e106811
14. Lammi L, Arte S, Somer M et al (2004) Muta- 27. Yamakoshi Y, Hu JC-C, Fukae M et al (2005)
tions in AXIN2 cause familial tooth agenesis Dentin glycoprotein the protein in the middle
and predispose to colorectal cancer. Am J of the dentin sialophosphoprotein chimera. J
Hum Genet 74(5):1043–1050 Biol Chem 280(17):17472–17479
15. Bergendal B, Klar J, Stecksén-Blicks C et al 28. Yang Q, Chen D, Xiong F et al (2016) A
(2011) Isolated oligodontia associated with splicing mutation in VPS4B causes dentin dys-
mutations in EDARADD, AXIN2, MSX1, plasia I. J Med Genet 53(9):624–633
and PAX9 genes. Am J Med Genet A 155 29. Xiong F, Ji Z, Liu Y et al (2017) Mutation in
(7):1616–1622 SSUH2 causes autosomal-dominant dentin
16. Arte S, Parmanen S, Pirinen S et al (2013) dysplasia type I. Hum Mutat 38(1):95–104
Candidate gene analysis of tooth agenesis iden- 30. Maxam AM, Gilbert W (1977) A new method
tifies novel mutations in six genes and suggests for sequencing DNA. Proc Natl Acad Sci 74
significant role for WNT and EDA signaling (2):560–564
and allele combinations. PLoS One 8(8): 31. Sanger F, Nicklen S, Coulson AR (1977) DNA
e73705 sequencing with chain-terminating inhibitors.
17. Crawford PJ, Aldred M, Bloch-Zupan A Proc Natl Acad Sci U S A 74(12):5463–5467
(2007) Amelogenesis imperfecta. Orphanet J 32. Mardis ER (2008) Next-generation DNA
Rare Dis 2(1):17 sequencing methods. Annu Rev Genomics
18. Smith CE, Murillo G, Brookes SJ et al (2016) Hum Genet 9:387–402
Deletion of amelotin exons 3–6 is associated 33. Smith M (2017) DNA sequence analysis in
with amelogenesis imperfecta. Hum Mol clinical medicine, proceeding cautiously. Front
Genet 25(16):3578–3587 Mol Biosci 4:24
19. Sillence D, Senn A, Danks D (1979) Genetic 34. Kircher M, Kelso J (2010) High-throughput
heterogeneity in osteogenesis imperfecta. J DNA sequencing—concepts and limitations.
Med Genet 16(2):101–116 BioEssays 32(6):524–536
20. Goldblatt J, Carman P, Sprague P (1991) 35. Koboldt DC, Larson DE, Chen K et al (2012)
Unique dwarfing, spondylometaphyseal skele- Massively parallel sequencing approaches for
tal dysplasia, with joint laxity and dentinogen- characterization of structural variation. Meth-
esis imperfecta. Am J Med Genet 39 ods Mol Biol 838:369–384
(2):170–172
36. Biesecker LG, Green RC (2014) Diagnostic
21. Unger S, Antoniazzi F, Brugnara M et al clinical genome and exome sequencing. N
(2008) Clinical and radiographic delineation Engl J Med 370(25):2418–2425
of odontochondrodysplasia. Am J Med Genet
A 146(6):770–778 37. Schouten JP, McElgunn CJ, Waaijer R et al
(2002) Relative quantification of 40 nucleic
22. Shields E, Bixler D, El-Kafrawy A (1973) A acid sequences by multiplex ligation-
proposed classification for heritable human dependent probe amplification. Nucleic Acids
dentine defects with a description of a new Res 30(12):e57–e57
entity. Arch Oral Biol 18(4):543–553, IN7
38. Schwartz M, Dunø M (2004) Improved
23. MacDougall M, Dong J, Acevedo AC (2006) molecular diagnosis of dystrophin gene muta-
Molecular basis of human dentin diseases. Am J tions using the multiplex ligation-dependent
Med Genet A 140(23):2536–2546 probe amplification method. Genet Test 8
24. de La Dure-Molla M, Fournier BP, Berdal A (4):361–367
(2015) Isolated dentinogenesis imperfecta and 39. Rhoads A, Au KF (2015) PacBio sequencing
dentin dysplasia: revision of the classification. and its applications. Genom Proteom Bioinfor-
Eur J Hum Genet 23(4):445–451 matics 13(5):278–289
25. MacDougall M, Simmons D, Luan X et al 40. O’Sullivan J, Bitu CC, Daly SB et al (2011)
(1997) Dentin phosphoprotein and dentin sia- Whole-Exome sequencing identifies FAM20A
loprotein are cleavage products expressed from mutations as a cause of amelogenesis imper-
a single transcript coded by a gene on human fecta and gingival hyperplasia syndrome. Am J
chromosome 4 Dentin phosphoprotein DNA Hum Genet 88(5):616–620
sequence determination. J Biol Chem 272
(2):835–842 41. Poulter JA, Brookes SJ, Shore RC et al (2014)
A missense mutation in ITGB6 causes pitted
26. Fisher LW, Fedarko NS (2003) Six genes hypomineralized amelogenesis imperfecta.
expressed in bones and teeth encode the cur- Hum Mol Genet 23(8):2189–2197
rent members of the SIBLING family of pro-
teins. Connect Tissue Res 44(1):33–40 42. Poulter JA, El-Sayed W, Shore RC et al (2014)
Whole-exome sequencing, without prior
linkage, identifies a mutation in LAMB3 as a 55. Stewart SK, Morris TJ, Guilhamon P et al
cause of dominant hypoplastic amelogenesis (2015) oxBS-450K: a method for analysing
imperfecta. Eur J Hum Genet 22(1):132–135 hydroxymethylation using 450K BeadChips.
43. Acevedo AC, Poulter JA, Alves PG et al (2015) Methods 72:9–15
Variability of systemic and oro-dental pheno- 56. Gross JA, Lefebvre F, Lutz P-E et al (2016)
type in two families with non-lethal Raine syn- Variations in 5-methylcytosine and
drome with FAM20C mutations. BMC Med 5-hydroxymethylcytosine among human
Genet 16(1):8 brain, blood, and saliva using oxBS and the
44. Yang J, Kawasaki K, Lee M et al (2016) The Infinium MethylationEPIC array. Biol Meth-
dentin phosphoprotein repeat region and ods Protocols 1(1):bpw002
inherited defects of dentin. Mol Genet Geno- 57. Field SF, Beraldi D, Bachman M et al (2015)
mic Med 4(1):28–38 Accurate measurement of 5-methylcytosine
45. Aidar M, Line SRP (2007) A simple and cost- and 5-hydroxymethylcytosine in human cere-
effective protocol for DNA isolation from buc- bellum DNA by oxidative bisulfite on an array
cal epithelial cells. Braz Dent J 18(2):148–152 (OxBS-array). PLoS One 10(2):e0118202
46. Lahiri DK, Nurnberger JI Jr (1991) A rapid 58. Skvortsova K, Zotenko E, Luu P-L et al (2017)
non-enzymatic method for the preparation of Comprehensive evaluation of genome-wide
HMW DNA from blood for RFLP studies. 5-hydroxymethylcytosine profiling approaches
Nucl Acids Res 19(19):5444 in human DNA. Epigenet Chromatin 10(1):16
47. Thiede C, Prange-Krex G, Freiberg-Richter J 59. Pastor WA, Pape UJ, Huang Y et al (2011)
et al (2000) Buccal swabs but not mouthwash Genome-wide mapping of
samples can be used to obtain pretransplant 5-hydroxymethylcytosine in embryonic stem
DNA fingerprints from recipients of allogeneic cells. Nature 473(7347):394–397
bone marrow transplants. Bone Marrow Trans- 60. Taiwo O, Wilson GA, Morris T et al (2012)
plantation 25(5):575 Methylome analysis using MeDIP-seq with low
48. Abraham JE, Maranian MJ, Spiteri I et al DNA concentrations. Nat Protoc 7
(2012) Saliva samples are a viable alternative (4):617–636
to blood samples as a source of DNA for high 61. Nestor CE, Meehan RR (2014) Hydroxy-
throughput genotyping. BMC Med Genomics methylated DNA immunoprecipitation (hme-
5(1):19 DIP). Func Anal DNA Chromatin
49. Ishmael FT, Stellato C (2008) Principles and 1094:259–267
applications of polymerase chain reaction: basic 62. Buenrostro JD, Wu B, Chang HY et al (2015)
science for the practicing physician. Ann ATAC-seq: a method for assaying chromatin
Allergy Asthma Immunol 101(4):437–443 accessibility genome-wide. Curr Protoc Mol
50. Kolmodin LA, Birch DE (2002) Polymerase Biol 109:21.29.21–21.29.29
chain reaction. Basic principles and routine 63. Chen S, Santos L, Wu Y et al (2005) Altered
practice. Methods Mol Biol 192:3–18 gene expression in human cleidocranial dyspla-
51. Untergasser A, Cutcutache I, Koressaar T et al sia dental pulp cells. Arch Oral Biol 50
(2012) Primer3—new capabilities and inter- (2):227–236
faces. Nucl Acids Res 40(15):e115 64. Yan W, Zhang C, Yang X et al (2015) Abnor-
52. Koressaar T, Remm M (2007) Enhancements mal differentiation of dental pulp cells in clei-
and modifications of primer design program docranial dysplasia. J Dent Res 94(4):577–583
Primer3. Bioinformatics 23(10):1289–1291 65. Mosmann T (1983) Rapid colorimetric assay
53. Landre P, Gelfand D, Watson R (1995) The for cellular growth and survival: application to
use of cosolvents to enhance amplification by proliferation and cytotoxicity assays. J Immu-
the polymerase chain reaction. In: PCR strate- nol Methods 65(1–2):55–63
gies. Academic, New York, pp 3–16 66. Liang C-C, Park AY, Guan J-L (2007) In vitro
54. Tang M, Xu W, Wang Q et al (2009) Potential scratch assay: a convenient and inexpensive
of DNMT and its epigenetic regulation for method for analysis of cell migration in vitro.
lung cancer therapy. Curr Genomics 10 Nat Protoc 2(2):329–333
(5):336–352
Chapter 38
Protocols, Methods, and Tools for Genome-Wide Association

Studies (GWAS) of Dental Traits
Cary S. Agler, Dmitry Shungin, Andrea G. Ferreira Zandoná,
Paige Schmadeke, Patricia V. Basta, Jason Luo, John Cantrell,
Thomas D. Pahel Jr., Beau D. Meyer, John R. Shaffer, Arne S. Schaefer,
Kari E. North, and Kimon Divaris
Abstract
Oral health and disease are known to be influenced by complex interactions between environmental (e.g.,
social and behavioral) factors and innate susceptibility. Although the exact contribution of genomics and
other layers of “omics” to oral health is an area of active research, it is well established that the susceptibility
to dental caries, periodontal disease, and other oral and craniofacial traits is substantially influenced by the
human genome. A comprehensive understanding of these genomic factors is necessary for the realization of
precision medicine in the oral health domain. To aid in this direction, the advent and increasing affordability
of high-throughput genotyping has enabled the simultaneous interrogation of millions of genetic poly-
morphisms for association with oral and craniofacial traits. Specifically, genome-wide association studies
(GWAS) of dental caries and periodontal disease have provided initial insights into novel loci and biological
processes plausibly implicated in these two common, complex, biofilm-mediated diseases. This paper
presents a summary of protocols, methods, tools, and pipelines for the conduct of GWAS of dental caries,
periodontal disease, and related traits. The protocol begins with the consideration of different traits for
both diseases and outlines procedures for genotyping, quality control, adjustment for population stratifica-
tion, heritability and association analyses, annotation, reporting, and interpretation. Methods and tools
available for GWAS are being constantly updated and improved; with this in mind, the presented
approaches have been successfully applied in numerous GWAS and meta-analyses among tens of thousands
of individuals, including dental traits such as dental caries and periodontal disease. As such, they can serve as
a guide or template for future genomic investigations of these and other traits.
Key words Genome-wide association studies, Methods, Bioinformatics, Dental caries, Periodontal
disease, Periodontitis, Protocol
1 Introduction
Oral health and disease endpoints are the result of complex inter-
actions between innate, behavioral, environmental, and social fac-
tors. An exhaustive model of risk and protective factors, as well as
493
494 Cary S. Agler et al.
behavioral and biologic factors at play, remains elusive and is,

arguably, unattainable. Nevertheless, advances in the oral and cra-
niofacial health sciences and a growing body of evidence have
illuminated the major influences on dental caries [1] and periodon-
tal disease [2]—the two main, common, complex oral diseases.
From a public health standpoint, both conditions are marked by
pronounced social gradients and disparities [3]. From a pathoge-
netic standpoint, they are both to some degree biofilm- and host
immunity-mediated [4–6] and modulated by behavior and host
genomics [7].
Meaningful improvements in the prevention and treatment of
oral diseases are likely to be achieved when aspects of precision
medicine [8, 9] are realized in the oral health domain, with geno-
mics being a key part of the puzzle. Although the exact contribu-
tion of genomics and other layers of “omics” to oral health is an
area of active investigation, it is well established that the suscepti-
bility to dental caries and periodontal disease is to some degree
determined by the human genome. Estimates of heritability for
dental caries and periodontitis reported in the literature vary sub-
stantially but are generally in the range of 20–50% [10–12], with
more severe or early-onset forms of disease being more heritable.
The recent advent and increasing affordability of high-
throughput genotyping have enabled comprehensive genomic
investigations of oral and craniofacial traits; however, the current
knowledge base of oral health genomics pales in comparison to
what is known for other human traits [13]. While major advances
have been made for several common diseases including asthma,
diabetes, cancer, and cardio-metabolic and psychiatric conditions,
genome-wide association studies (GWAS) of dental caries [14–17]
and periodontal diseases [18–28] have provided only initial insights
into plausible novel loci and biological processes implicated in these
two common, complex, biofilm-mediated diseases.
To aid in the conduct and harmonization of genomic investiga-
tions in the field and offer an overview of current common ground
procedures, this paper presents a summary of protocols, methods,
tools, and pipelines for the conduct of GWAS of dental caries,
periodontal diseases, and related traits. First, it describes various
traits for both diseases that can be carried forward to GWAS.
Further, it outlines the major steps involved in genotyping, impu-
tation, quality control, adjustment for population stratification,
heritability and association analyses, annotation, reporting, and
interpretation. Methods and tools available for GWAS are being
constantly updated and improved—with this in mind, the
approaches presented in this paper have been successfully applied
in numerous GWAS and meta-analyses among tens of thousands of
individuals, including dental traits such as dental caries and peri-
odontal disease. As such, they can serve as a guide or template for
future genomic investigations of these and other traits.
Methods for GWAS of Dental Traits 495
2 Materials
2.1 Sources of 1. Clinical examinations conducted by trained, ideally calibrated

Phenotype Information examiners; an examiner is typically considered calibrated upon
achievement of preset levels of inter-examiner and intra-
examiner agreement, usually based on weighted or unweighted
kappas, percent agreement, or other criteria.
2. Clinical records (e.g., electronic patient record data).
3. Intraoral photographs, scored by trained and calibrated evalua-
tors or algorithms/software.
4. Administrative claims data (e.g., inferred by dental caries or
periodontitis-related treatment procedures).
5. Self-reported data (e.g., obtained via written or electronic
questionnaires or via telephone).
2.2 Sources of DNA 1. Blood samples.

for Genotyping 1. Saliva samples [e.g., DNA Genotek Oragene DNA (OG)—500
kit for adults or OG 575 kit for young children or non-spitters,
DNA Genotek, Ottawa, Ontario, Canada].
2. Buccal swabs (e.g., MAWI technologies iSWAB™ kit, Mawi
DNA Technologies LLC, Hayward, CA).
3. Newborn blood spots [29].
2.3 DNA Extraction, 1. Automated DNA extraction from whole blood with either
Quantitation, and high-salt extraction methods or automated magnetic-bead
Quality Assessment extraction methods (i.e., PerkinElmer® Chemagic™ MSM I
robotic system) or manual DNA extraction methods.
2. Automated DNA extraction from saliva, buccal brushes, or
blood spots with automated magnetic-bead extraction meth-
ods (i.e., PerkinElmer® Chemagic™ MSM I robotic system) or
manual DNA extraction methods.
3. DNA quantitation using NanoDrop™ spectrophotometry,
Quant-iT™ PicoGreen® fluorometry, Qubit™ fluorometry,
or human-specific RNaseP assays.
4. Quality (i.e., sample purity) assessment using spectrophoto-
metric methods (e.g., 260:280 and 260:230 absorbance
ratios).
2.4 Genotyping 1. High-density genotyping arrays (e.g., Illumina Infinium

Supplies, Equipment, Omni5Exome—4 BeadChip array, offering ~4.3 million var-
and Software iants and exome content) or targeted genotyping arrays (e.g.,
Illumina Metabochip or Immunochip).
2. Array scanning (e.g., Illumina iScan).
3. Variant calling software (e.g., Illumina GenomeStudio).
2.5 Imputation As referenced below in Subheading 3.3, step 4, currently the

Software University of Michigan Imputation Server is most frequently used
for this step:
1. Eagle2 [30].
2. MiniMac [31].
3. IMPUTE2 [32].
4. GeneImp [33].
2.6 Genotype 1. Cloud- or intranet-based storage with secure File Transfer

Storage, Transfer, Protocol (FTP) capabilities.
and Management 2. High-performance computing cluster allowing multi-
threading and large memory jobs.
3. Server or workstations with common data management and
programming suites (e.g., R, SAS, Stata).
2.7 Software 2. PLINK [34] most commonly used.

Commonly Used for 3. SUGEN [35] implements weighted estimators to account for
Genome-Wide unequal sampling probability in complex survey design
Association Analyses settings.
4. SNPTEST [36] analysis of single SNP frequentist and Bayesian
association tests.
5. GenABEL Project suite [37, 38] GWAS analyses and statistical
“omics” applications, including mixed model-based GWAS.
6. GWASTools [39], an R/Bioconductor package for quality
control and association analyses.
7. EMMAX [40], SOLAR [41, 42], GEMMA [43] mixed models
for association analysis accounting for sample structure (e.g.,
family-based studies).
8. GMMAT [44] implemented via GENESIS [45] generalized
linear mixed model test for GWAS of binary traits accounting
for population structure and relatedness, available via the
GENESIS R/Bioconductor package.
2.8 Software Used 1. GCTA [46], heritability estimation and generation of eigen-
for Supporting vectors for ancestry adjustment.
Analysis Routines and 2. EIGENSTRAT [47], principal component analysis-based cor-
Visualizations rection for population stratification.
3. ADMIXTURE [48], model-based estimation of ancestry in
unrelated individuals.
4. MAGENTA [49], gene-centric and pathway/gene-set enrich-
ment analyses.
5. MAGMA [50], gene-set analysis of GWAS data.
6. VEGAS/VEGAS2 [51], versatile gene-based association

analysis.
7. LocusZoom [52], local or web-based creation of regional asso-
ciation plots.
2.9 Software Used 1. METAL [53], meta-analysis of GWAS results.

for Meta-analysis and 2. GWAMA [54], meta-analysis of GWAS summary statistics.
Subsequent Quality
3. EasyStrata [55], evaluation and visualization of stratified
Control and Post-
GWAS meta-analysis data.
processing of GWAS
Results
4. SQC [56], secure quality control for GWAS meta-analyses.
2.10 Resources for 1. ENCODE Project [57].

Genomic Context and 2. Roadmap Epigenomics Project [58].
Functional
3. UCSC Genome Browser [59].
Annotations
4. Integrative Genomics Viewer (IGV) [60].
5. GTEx [61], genotype-tissue expression project.
6. LDlink [62], exploration of population-specific haplotype
structures and links the alleles of interest with possible func-
tional variants.
7. PolyPhen-2 [63], prediction of functional effects of human
non-synonymous coding single nucleotide polymorphisms
(SNPs).
8. ScanDB [64], database including physical- and function-based
annotation of SNPs (e.g., association with gene expression).
9. Genehopper [65], multidimensional scoring of gene-gene
interactions.
3 Methods
Obtaining high-quality phenotypes is the first important step

before the actual conduct of a GWAS. Similar to any other type of
study, traits carried forward to GWAS are subject to systematic and
random errors, which are threats to the validity and precision of the
obtained results. Arguably, reliance on very large sample sizes (tens
or hundreds of thousands of individuals) may allow for the detec-
tion of true genetic signals despite some trait misclassification;
however, in principle, higher-quality phenotypes (e.g., clinical or
biological) provide “better” results compared to lesser-quality traits
(e.g., self-reported) for a given sample size. Numerous traits are
available for the study of dental caries and periodontal disease in the
context of a GWAS—Table 1 presents a non-exhaustive list of such
traits for dental caries (including early childhood caries or ECC
Table 1
Example traits for dental caries, developmental defects of the enamel, periodontal disease, and
edentulism that can be interrogated in the context of GWAS
Condition Trait Type Description

Dental caries DMFS/dmfs Continuous The sum of decayed, missing due to caries,
filled/restored tooth surfaces; five
surfaces are enumerated for molars (and
premolars, in the permanent dentition),
four for all other teeth [14]
DMFS/dmfs > 0 Binary Case definition for caries experience vs.
caries-free; this definition corresponds to
“early childhood caries” or ECC among
children <72 months of age
[17, 66]. defs and dfs indices may also be
considered in the primary dentition
DFS/S Continuous/ The proportion of diseased tooth surfaces
proportion among the surfaces present [14]
DMTFS Continuous The sum of decayed, missing due to all
reasons (presumably mainly caries and
periodontal disease) and filled/restored
surfaces. A tooth morbidity index [91]
Developmental DDE > 0 Binary The presence of any developmental defect
defects of the of the enamel, as defined by the Clarkson
enamel and O’Mullane modified epidemiologic
index [67], of diameter 1 mm of greater,
as assessed on the facial or labial surface
of the entire dentition
Periodontal Chronic Categorical Based upon probing depth (PD) and
disease periodontitis, clinical attachment loss (AL), categories
CDC/AAP of health/no disease, mild, moderate,
definition (2012) and severe disease are defined
[92]. Severe disease is defined as
2 interproximal sites with AL 6 mm
(not on same tooth) and 1
interproximal site with PD
5 mm [27, 92]
Periodontal extent Continuous/ Proportion of all examined sites with
scores proportion probing depth, attachment loss equal or
greater to a predefined threshold (e.g.,
3 mm, 4 mm, etc.), or bleeding upon
probing, e.g., extent of severe gingival
bleeding [21]
Full-mouth summary Continuous Summary scores or means of periodontal
scores indices, e.g., mean interproximal clinical
attachment level [19]
UNC periodontal Categorical Latent class analysis-based categories of
profile class (PPC) periodontal disease experience
system incorporating patterns of tooth loss [85]
(continued)
Table 1
(continued)
Condition Trait Type Description

Periodontal complex Principal Complex periodontal traits derived by
traits (PCTs) component- principal component analysis of clinical,
derived inflammatory (e.g., gingival crevicular
eigenvectors fluid IL–1b), and microbial (e.g.,
periodontal pathogen levels) data [20]
Aggressive Binary Radiographically confirmed 50% bone
periodontitis loss at two to six teeth, diagnosed at age
(AgP), localized of 35 or younger [93]
Aggressive Binary Radiographically confirmed 50% bone
periodontitis loss at 7 teeth, diagnosed at age 35 or
(AgP), generalized younger [93]
Edentulism Number of remaining Continuous The sum of remaining natural teeth (0–28
natural teeth or 0–32, depending on the inclusion of
third molars)
“Functional Binary The World Health Organization definition
dentition” of reduced dental arches that preserve
basic functions—i.e., the retention of a
natural, esthetic, functional dentition of
no less than 20 teeth throughout life
with no need for tooth replacement [86]
No remaining natural Binary Complete loss of the natural dentition
teeth
[66]), developmental defects of the enamel (DDE) [67], periodon-

tal disease, and edentulism phenotype definitions.
As a second step, it is imperative that investigators have a good
understanding of the study design under consideration and the
characteristics of the sample; these factors may have an influential
impact on the obtained results. For example, case-control studies
and small sample sizes are known to produce spurious results due to
systematic (e.g., Berksonian [68]) bias and random error. Third,
application of strict criteria for determining what is a significant
signal or variant is a requirement from a scientific rigor standpoint
[69]. Two-stage designs and discovery-and-replication approaches
are common strategies used to reduce the unavoidable type I errors
in the single-stage or discovery samples and reduce the “winner’s
curse” phenomenon [70]. Finally, an additional consideration prior
to the conduct of a GWAS is the availability of sufficient quantity
and quality of DNA. This is usually not an issue, unless a pediatric
population is under study—young children are expected to be less
cooperative than adolescents or adults with research procedures
involving venipuncture, as well as produce less saliva. With these
in mind, carry out the procedures as follows:
3.1 DNA Extraction 1. Extract, quantify, and quality-assess DNA from blood, saliva, or
buccal swab samples, according to the extraction kit manufac-
turer’s instructions.
2. At least 400 ng of DNA is required for genotyping with
the Infinium Omni5Exome-4 BeadChip referenced in Sub-
heading 2.4.
3. Plate extracted DNA, and ship to genotyping core using a
manifest according to each facility’s instructions and best prac-
tices for safety and sample integrity.
3.2 Genotype Quality 1. An excellent, comprehensive description of quality control

Control, Variant procedures is available in the Supplementary Material accom-
Calling, and panying the 2007 Wellcome Trust GWAS [36].
Exclusions 2. Use manufacturer’s instructions for processing and scanning of
samples.
3. Use HapMap-CEPH trios and duplicates and blind duplicate
samples for quality control.
4. Use GenomeStudio (https://goo.gl/BmNin1) or any other
software for genotype calling [e.g., SNPs and copy number
variants (CNV)].
5. Generate sample call and error rates.
6. Identify sex mismatches and relationship errors (annotated
versus genetic), gross chromosomal anomalies, mosaicism,
contamination, and sample swaps.
7. Discard individual samples and markers that fail quality control
criteria—these are highly dependent on the study design and
sample (e.g., ranges of between 90 and 99% completion for
markers and samples, duplicate sample discordances, Mende-
lian errors).
8. Deposit/upload SNP genotype data for transfer to imputation
and association analyses; raw data or intensity files may be used
for variant calling with third party software or uploaded for
long-term storage.
3.3 Imputation of 1. Use TOPMed, HapMap, 1000 Genomes Project, Consortium

Genotypes on Asthma among African-ancestry Populations in the Amer-
icas (CAAPA), or the Haplotype Reference Consortium
(HRC) reference panels (including multiple other cohorts)
for imputation.
2. Pre-phase (estimate haplotypes) using SHAPEIT [71, 72], as
needed.
3. Use any available imputation software pipeline, Eagle2, Mini-
Mac, IMPUTE2, or GeneImp, referenced in Subheading 2.5,
as needed.
4. Use these programs on premise or on the cloud; e.g., the

University of Michigan Imputation Server [73] provides free
genotype imputation services using Minimac3.
5. Apply quality filters to imputed SNPs and exclude based upon
any of several available criteria [74] [e.g., based upon thresh-
olds of imputation quality score, expected r2 or “info” score,
Pearson correlation between true and “best-guess” genotypes
(R2), high missing data rate after imputation, or other available
metrics].
6. Further exclude rare [minor allele frequency (MAF) <5%,
<1%, or any other preset threshold] and monomorphic SNPs.
3.4 Adjustment for 1. Create eigenvectors or adjust for admixture proportion for
Ancestry and population stratification control using programs referenced in
Population Subheading 2.8.
Stratification 2. Alternatively use mixed model-based approaches using tools
referenced in Subheading 2.8 to model putative genetic
clustering.
3. Exclude genetic outliers or related individuals (e.g., first-
and/or second-degree individuals) if the study design assumes
a sample of unrelated individuals.
3.5 GWAS Analysis 1. Prepare the phenotype data using appropriate transformations,
as necessary and dependent on the trait distribution
characteristics.
2. Account for population stratification using eigenvectors or
admixture proportions developed as described in Subheading
3.4, as well as other study and participants’ characteristics—
typically sex, age (with or without an age-squared term), exam-
ination center, as well as complex survey design or familial
relatedness (using software [40–45] detailed in Subheading
2.7), as necessary.
3. Use appropriate statistical models to test variant-phenotype
associations—commonly used methods include linear regres-
sion modeling for continuous traits after appropriate transfor-
mation if needed (e.g., DMFS index) and logistic regression for
binary outcomes (e.g., ECC case status and chronic
periodontitis).
4. Consider a P value of less than 5 10 8 as the conventional
genome-wide statistical significance threshold for GWAS,
assuming one million independent association tests and a Bon-
ferroni multiple testing correction.
5. One can carry out stratified GWAS analyses, on any variable of
interest, depending on the research question (e.g., strata of sex,
ancestry, or any strong risk factor such as fluoride for dental
caries and smoking or diabetes for periodontal disease) and the

study design.
6. Combine results, if applicable, using software and methods
detailed in Subheading 2.8, using weights proportional to the
square root of each study’s sample size (i.e., Stouffer’s method
[75]) or by each study estimates’ standard errors.
7. For trans-ethnic meta-analyses, one can use one of several
additional approaches for meta-analyses based on random or
fixed effects (e.g., trans-ethnic meta-analyses), as well as
two-stage designs [76–80].
8. Generate QQ plots of association, and compute the genomic
inflation factor (lambda) to examine for possible systematic
departures from expectation due to residual population strati-
fication or other sources.
9. Generate Manhattan plots to visualize genome-wide results
and genetic signals.
3.6 Reporting and 1. For each SNP, it is customary to report key information (e.g.,
Annotation chromosome position, reference genome build, strand, refer-
ence allele, minor allele frequency, n, beta, standard error, 95%
confidence interval, p-value, genotyped or imputed indicator,
imputation quality score, as well as meta-analysis statistics (e.g.,
effect size with corresponding uncertainly estimate and p-
values) if applicable.
2. Provide annotation for genomic context, linkage disequilib-
rium with functional variants, relative position or distance
from exons, or exon boundaries and gene promoter regions.
3. Investigate the predicted effect of non-synonymous coding
variants on the protein structure and function using
PolyPhen-2 [63] as referenced in Subheading 2.10.
4. Create regional association plots using LocusZoom [52] as
referenced in Subheading 2.8.
5. Visualize, compare, and contrast results with other GWAS
findings of the same, similar, or potentially related traits using
IGV [60] as referenced in Subheading 2.10.
3.7 Post-GWAS 1. Conduct gene-centric and pathway or gene-set enrichment

Analyses and analyses using MAGENTA [49], MAGMA [50], or other
Procedures tools referenced in Subheading 2.8.
2. Examine individual marker associations with functional
genome elements and gene expression using software or
resources detailed in Subheading 2.10.
3. Typically share or upload GWAS results onto a study-specific or
“community” server, to enable harmonization and meta-
analyses or dissemination of results.
4 Notes
1. Although here we include DNA extraction, genotyping, and

associated quality control procedures, most groups collaborate
with core facilities to carry out these steps to produce
genotype data.
2. Several steps of this protocol can be slow or time-consuming—
for instance, genotyping of a large cohort can take months.
3. SNPs identified as genome-wide significant are unlikely to be
causal variants but rather highlight genome areas (loci) of
interest, where causative variants may lie.
4. Most GWAS signals are located in non--coding areas, where
regulatory functional [57] elements of the genome are
enriched. Variants in regulatory DNA sites may have distant
gene targets, and have been shown to systematically affect
transcription factor recognition sequences, alter chromatin
states and form regulatory networks [81].
5. The interrogation of disease-associated variants’ correlation
with relevant tissue-specific gene expression is a promising
follow-up strategy of GWAS signals.
6. GWAS offer improvements over candidate-gene studies, but if
sample sizes are small, they are still prone to type I and type II
errors—most signals identified in well-powered studies subse-
quently replicate.
7. Replicate novel signals based on directional consistency, nomi-
nal or (ideally) genome-wide significant association in external,
independent samples.
8. The combination of multiple cohorts with phenotype and
genotype data to interrogate oral and craniofacial health traits
(e.g., caries and periodontitis) is warranted—dental consortia
have begun forming (e.g., the GLIDE [82]). Despite some
unavoidable challenges with achieving phenotype harmoniza-
tion and addressing differences between populations (e.g., var-
iations in disease prevalence between North American and
European studies), single cohort reports are less informative
and may be inefficient if realistic opportunities for joint, collab-
orative analyses exist.
9. Whenever feasible, follow up promising, replicated loci with
experimental (e.g., animal) models and other systems biology
approaches.
10. Arguably, biologically informed traits [20], data-driven clusters
[83, 84], or endophenotypes have the potential to be informa-
tive targets for interrogation in the GWAS context, especially in
cases where clinical endpoints are marked by heterogeneity and

are substantially modified by dental treatment.
11. Third molars are frequently excluded from dental caries and
periodontitis indices and definitions—this should be made
explicit and justified in each cohort/study.
12. Tooth loss influences epidemiologic measures of periodontal
disease (e.g., most periodontally affected teeth may be lost).
For this reason, explicit consideration of tooth loss patterns
[85] can offer advantages when interrogating cross-sectional
datasets of periodontitis endpoints, by capturing individuals
that may otherwise be misclassified. Similarly, interrogation of
edentulism [86] and dental caries experience patterns can be a
promising strategy for dental caries GWAS, both in the primary
[84, 87] and the permanent [83, 88, 89] dentition.
13. Two-step approaches including both genotyping and WGS in a
subset of a cohort are considered efficient and cost-effective
under certain circumstances to boost power to detect rare
variants. For example, selection of a subset (e.g., 10%) of the
‘most diseased’ participants in the sample for WGS and their
subsequent use to enrich the imputation reference panels in
integrative analyses, assuming that they are more likely to
harbor rare, causal variants), has been recently proposed as an
efficient strategy [90].
Acknowledgments
This work was supported by grants from the National Institutes of

Health, National Institute of Dental and Craniofacial Research,
U01-DE025046, R56-027055 and R03-DE024264, and a grant
from the Swedish Research Council 4.1-2016-00416.
References
1. Selwitz RH, Ismail AI, Pitts NB (2007) Dental 6. Hajishengallis G (2014) Immunomicrobial
caries. Lancet 369(9555):51–59 pathogenesis of periodontitis: keystones,
2. Offenbacher S (1996) Periodontal diseases: pathobionts, and host response. Trends Immu-
pathogenesis. Ann Periodontol 1(1):821–878 nol 35(1):3–11
3. Lee JY, Divaris K (2014) The ethical imperative 7. Nibali L, Di Iorio A, Tu YK, Vieira AR (2017)
of addressing oral health disparities: a unifying Host genetics role in the pathogenesis of peri-
framework. J Dent Res 93(3):224–230 odontal disease and caries. J Clin Periodontol
4. Nyvad B, Crielaard W, Mira A, Takahashi N, 44 Suppl 18:S52–S78
Beighton D (2013) Dental caries from a 8. Collins FS, Varmus H (2015) A new initiative
molecular microbiological perspective. Caries on precision medicine. N Engl J Med 372
Res 47(2):89–102 (9):793–795
5. Genco RJ, Borgnakke WS (2013) Risk factors 9. Divaris K (2017) Fundamentals of precision
for periodontal disease. Periodontology 62 medicine. Compend Contin Educ Dent 38
(1):59–94 Suppl 8:30–32
10. Vieira AR, Modesto A, Marazita ML (2014) Saruhan-Direskeneli G, Guzeldemir-

Caries: review of human genetics research. Car- Akcakanat E, Keceli G, Laudes M, Noack B,
ies Res 48(5):491–506 Teumer A, Holtfreter B, Kocher T, Eickholz P,
11. Michalowicz BS, Diehl SR, Gunsolley JC, Meyle J, Doerfer C, Bruckmann C, Lieb W,
Sparks BS, Brooks CN, Koertge TE, Califano Franke A, Schreiber S, Nohutcu RM,
JV, Burmeister JA, Schenkein HA (2000) Evi- Erdmann J, Loos BG, Jepsen S,
dence of a substantial genetic basis for risk of Dommisch H, Schaefer AS (2017) A genome-
adult periodontitis. J Periodontol 71 wide association study identifies nucleotide var-
(11):1699–1707 iants at SIGLEC5 and DEFA1A3 as risk loci for
12. Loos BG, Papantonopoulos G, Jepsen S, Laine periodontitis. Hum Mol Genet 26
ML (2015) What is the contribution of genet- (13):2577–2588
ics to periodontal risk? Dent Clin N Am 59 19. Sanders AE, Sofer T, Wong Q, Kerr KF,
(4):761–780 Agler C, Shaffer JR, Beck JD, Offenbacher S,
13. Welter D, MacArthur J, Morales J, Burdett T, Salazar CR, North KE, Marazita ML, Laurie
Hall P, Junkins H, Klemm A, Flicek P, CC, Singer RH, Cai J, Finlayson TL, Divaris K
Manolio T, Hindorff L, Parkinson H (2014) (2017 Jan) Chronic Periodontitis Genome-
The NHGRI GWAS Catalog, a curated wide Association Study in the Hispanic Com-
resource of SNP-trait associations. Nucleic munity Health Study/Study of Latinos. J Dent
Acids Res 42(Database issue):D1001–D1006 Res 96(1):64–72
14. Wang X, Shaffer JR, Zeng Z, Begum F, Vieira 20. Offenbacher S, Divaris K, Barros SP, Moss KL,
AR, Noel J, Anjomshoaa I, Cuenco KT, Lee Marchesan JT, Morelli T, Zhang S, Kim S,
MK, Beck J, Boerwinkle E, Cornelis MC, Hu Sun L, Beck JD, Laudes M, Munz M, Schaefer
FB, Crosslin DR, Laurie CC, Nelson SC, AS, North KE (2016) Genome-wide associa-
Doheny KF, Pugh EW, Polk DE, Weyant RJ, tion study of biologically informed periodontal
Crout R, McNeil DW, Weeks DE, Feingold E, complex traits offers novel insights into the
Marazita ML (2012) Genome-wide association genetic basis of periodontal disease. Hum Mol
scan of dental caries in the permanent denti- Genet 25(10):2113–2129
tion. BMC Oral Health 12:57 21. Zhang S, Divaris K, Moss K, Yu N, Barros S,
15. Sofer T, Shaffer JR, Graff M, Qi Q, Stilp AM, Marchesan J, Morelli T, Agler C, Kim SJ,
Gogarten SM, North KE, Isasi CR, Laurie CC, Wu D, North KE, Beck J, Offenbacher S
Szpiro AA (2016) Meta-analysis of Genome- (2016) The novel ASIC2 locus is associated
Wide Association studies with correlated indi- with severe gingival inflammation. JDR Clin
viduals: application to the Hispanic Commu- Trans Res 1(2):163–170
nity Health Study/Study of Latinos (HCHS/ 22. Shimizu S, Momozawa Y, Takahashi A,
SOL). Genet Epidemiol 40(6):492–501 Nagasawa T, Ashikawa K, Terada Y, Izumi Y,
16. Shaffer JR, Wang X, Feingold E, Lee M, Kobayashi H, Tsuji M, Kubo M, Furuichi Y
Begum F, Weeks DE, Cuenco KT, Barmada (2015) A genome-wide association study of
MM, Wendell SK, Crosslin DR, Laurie CC, periodontitis in a Japanese population. J Dent
Doheny KF, Pugh EW, Zhang Q, Feenstra B, Res 94(4):555–561
Geller F, Boyd HA, Zhang H, Melbye M, Mur- 23. Hong KW, Shin MS, Ahn YB, Lee HJ, Kim HD
ray JC, Weyant RJ, Crout R, McNeil DW, Levy (2015) Genomewide association study on
SM, Slayton RL, Willing MC, Broffitt B, Vieira chronic periodontitis in Korean population:
AR, Marazita ML (2011) Genome-wide asso- results from the Yangpyeong health cohort. J
ciation scan for childhood caries implicates Clin Periodontol 42:703. https://doi.org/10.
novel genes. J Dent Res 90(12):1457–1462 1111/jcpe.12437.
17. Ballantine J, Carlson JC, Zandona A, Agler CS, 24. Feng P, Wang X, Casado PL, Küchler EC,
Zeldin LP, Rozier G, Roberts MW, Basta PV, Deeley K, Noel J, Kimm H, Kim JH, Haas
Luo J, Antonio-Obese ME, McNeil D, AN, Quinelato V, Bonato LL, Granjeiro JM,
Weyant R, Crout RJ, Slayton R, Levy S, Shaffer Susin C, Vieira AR (2014) Genome wide asso-
JR, Marazita ML, North KE, Divaris K (2018) ciation scan for chronic periodontitis implicates
Exploring the genomic basis of early childhood novel locus. BMC Oral Health 14:84
caries: a pilot study. Int J Paediatr Dent 28 25. Shaffer JR, Polk DE, Wang X, Feingold E,
(2):217–225 Weeks DE, Lee MK, Cuenco KT, Weyant RJ,
18. Munz M, Willenborg C, Richter GM, Jockel- Crout RJ, McNeil DW, Marazita ML (2014)
Schneider Y, Graetz C, Staufenbiel I, Genome-wide association study of periodontal
Wellmann J, Berger K, Krone B, Hoffmann P, health measured by probing depth in adults
van der Velde N, Uitterlinden AG, de Groot ages 18–49 years. G3 (Bethesda) 4
LCPGM, Sawalha AH, Direskeneli H, (2):307–314
26. Teumer A, Holtfreter B, Völker U, Study of Latinos. Am J Hum Genet 95

Petersmann A, Nauck M, Biffar R, Völzke H, (6):675–688
Kroemer HK, Meisel P, Homuth G, Kocher T 36. Wellcome Trust Case Control Consortium
(2013) Genome-wide association study of (2007) Genome-wide association study of
chronic periodontitis in a general German pop- 14,000 cases of seven common diseases and
ulation. J Clin Periodontol 40(11):977–985 3,000 shared controls. Nature 447
27. Divaris K, Monda KL, North KE, Olshan AF, (7145):661–678
Reynolds LM, Hsueh WC, Lange EM, Moss K, 37. Karssen LC, van Duijn CM, Aulchenko YS
Barros SP, Weyant RJ, Liu Y, Newman AB, (2016) The GenABEL Project for statistical
Beck JD, Offenbacher S (2013) Exploring the genomics. F1000Res 5:914
genetic basis of chronic periodontitis: a 38. Aulchenko YS, Ripke S, Isaacs A, van Duijn
genome-wide association study. Hum Mol CM (2007) GenABEL: an R library for
Genet 22(11):2312–2324 genome-wide association analysis. Bioinfor-
28. Schaefer AS, Richter GM, Nothnagel M, matics 23(10):1294–1296
Manke T, Dommisch H, Jacobs G, Arlt A, 39. Gogarten SM, Bhangale T, Conomos MP,
Rosenstiel P, Noack B, Groessner-Schreiber B, Laurie CA, McHugh CP, Painter I, Zheng X,
Jepsen S, Loos BG, Schreiber S (2010) A Crosslin DR, Levine D, Lumley T, Nelson SC,
genome-wide association study identifies Rice K, Shen J, Swarnkar R, Weir BS, Laurie
GLT6D1 as a susceptibility locus for periodon- CC (2012) GWASTools: an R/Bioconductor
titis. Hum Mol Genet 19(3):553–562 package for quality control and analysis of
29. St Julien KR, Jelliffe-Pawlowski LL, Shaw GM, genome-wide association studies. Bioinformat-
Stevenson DK, O’Brodovich HM, Krasnow ics 28(24):3329–3331
MA, Stanford BPD Study Group (2013) 40. Kang HM, Sul JH, Service SK, Zaitlen NA,
High quality genome-wide genotyping from Kong SY, Freimer NB, Sabatti C, Eskin E
archived dried blood spots without DNA (2010) Variance component model to account
amplification. PLoS One 8(5):e64710 for sample structure in genome-wide associa-
30. Loh PR, Danecek P, Palamara PF, tion studies. Nat Genet 42(4):348–354
Fuchsberger C, A Reshef Y, K Finucane H, 41. Fardo DW, Zhang X, Ding L, He H,
Schoenherr S, Forer L, McCarthy S, Abecasis Kurowski B, Alexander ES, Mersha TB,
GR, Durbin R, L Price A (2016) Reference- Pilipenko V, Kottyan L, Nandakumar K, Mar-
based phasing using the Haplotype Reference tin L (2014) On family-based genome-wide
Consortium panel. Nat Genet 48 association studies with large pedigrees: obser-
(11):1443–1448 vations and recommendations. BMC Proc 8
31. Browning BL, Browning SR (2016) Genotype (Suppl 1 Genetic Analysis Workshop 18Vanessa
imputation with millions of reference samples. Olmo):S26
Am J Hum Genet 98(1):116–126 42. Almasy L, Blangero J (1998) Multipoint
32. Howie B, Fuchsberger C, Stephens M, March- quantitative-trait linkage analysis in general
ini J, Abecasis GR (2012) Fast and accurate pedigrees. Am J Hum Genet 62(5):1198–1211
genotype imputation in genome-wide associa- 43. Zhou X, Stephens M (2012) Genome-wide
tion studies through pre-phasing. Nat Genet efficient mixed-model analysis for association
44(8):955–959 studies. Nat Genet 44(7):821–824
33. Spiliopoulou A, Colombo M, Orchard P, 44. Chen H, Wang C, Conomos MP, Stilp AM,
Agakov F, McKeigue P (2017) GeneImp: fast Li Z, Sofer T, Szpiro AA, Chen W, Brehm
imputation to large reference panels using JM, Celedón JC, Redline S, Papanicolaou GJ,
genotype likelihoods from ultralow coverage Thornton TA, Laurie CC, Rice K, Lin X
sequencing. Genetics 206(1):91–104 (2016) Control for population structure and
34. Purcell S, Neale B, Todd-Brown K, Thomas L, relatedness for binary traits in genetic associa-
Ferreira MA, Bender D, Maller J, Sklar P, de tion studies via logistic mixed models. Am J
Bakker PI, Daly MJ, Sham PC (2007) PLINK: Hum Genet 98(4):653–666
a tool set for whole-genome association and 45. Conomos MP, Thornton T, Gogarten SM
population-based linkage analyses. Am J Hum (2017) GENESIS: GENetic EStimation and
Genet 81(3):559–575 Inference in Structured samples (GENESIS):
35. Lin DY, Tao R, Kalsbeek WD, Zeng D, Gon- statistical methods for analyzing genetic data
zalez F 2nd, Fernández-Rhodes L, Graff M, from samples with population structure
Koch GG, North KE, Heiss G (2014) Genetic and/or relatedness. R package version 2.6.1
association analysis under complex survey sam- 46. Yang J, Lee SH, Goddard ME, Visscher PM
pling: the Hispanic Community Health Study/ (2011) GCTA: a tool for genome-wide
complex trait analysis. Am J Hum Genet 88 Lister R, Hong C, Gascard P, Mungall AJ,
(1):76–82 Moore R, Chuah E, Tam A, Canfield TK, Han-
47. Price AL, Patterson NJ, Plenge RM, Weinblatt sen RS, Kaul R, Sabo PJ, Bansal MS, Carles A,
ME, Shadick NA, Reich D (2006) Principal Dixon JR, Farh KH, Feizi S, Karlic R, Kim AR,
components analysis corrects for stratification Kulkarni A, Li D, Lowdon R, Elliott G, Mercer
in genome-wide association studies. Nat Genet TR, Neph SJ, Onuchic V, Polak P,
38(8):904–909 Rajagopal N, Ray P, Sallari RC, Siebenthall
48. Alexander DH, Novembre J, Lange K (2009) KT, Sinnott-Armstrong NA, Stevens M, Thur-
Fast model-based estimation of ancestry in man RE, Wu J, Zhang B, Zhou X, Beaudet AE,
unrelated individuals. Genome Res 19 Boyer LA, De Jager PL, Farnham PJ, Fisher SJ,
(9):1655–1664 Haussler D, Jones SJ, Li W, Marra MA, McMa-
nus MT, Sunyaev S, Thomson JA, Tlsty TD,
49. Segrè AV, Wei N, DIAGRAM Consortium; Tsai LH, Wang W, Waterland RA, Zhang MQ,
MAGIC Investigators, Altshuler D, Florez JC Chadwick LH, Bernstein BE, Costello JF,
(2015) Pathways targeted by antidiabetes Ecker JR, Hirst M, Meissner A,
drugs are enriched for multiple genes asso- Milosavljevic A, Ren B, Stamatoyannopoulos
ciated with type 2 diabetes risk. Diabetes 64 JA, Wang T, Kellis M (2015) Integrative analy-
(4):1470–1483 sis of 111 reference human epigenomes.
50. de Leeuw CA, Mooij JM, Heskes T, Posthuma Nature 19;518(7539):317–330
D (2015) MAGMA: generalized gene-set anal- 59. Tyner C, Barber GP, Casper J, Clawson H,
ysis of GWAS data. PLoS Comput Biol 11(4): Diekhans M, Eisenhart C, Fischer CM,
e1004219 Gibson D, Gonzalez JN, Guruvadoo L,
51. Mishra A, Macgregor S (2015) VEGAS2: soft- Haeussler M, Heitner S, Hinrichs AS,
ware for more flexible gene-based testing. Twin Karolchik D, Lee BT, Lee CM, Nejad P,
Res Hum Genet 18(1):86–91 Raney BJ, Rosenbloom KR, Speir ML,
52. Pruim RJ, Welch RP, Sanna S, Teslovich TM, Villarreal C, Vivian J, Zweig AS, Haussler D,
Chines PS, Gliedt TP, Boehnke M, Abecasis Kuhn RM, Kent WJ (2017) The UCSC
GR, Willer CJ (2010) LocusZoom: regional Genome Browser database: 2017 update.
visualization of genome-wide association scan Nucleic Acids Res 45(D1):D626–D634
results. Bioinformatics 26(18):2336–2337 60. Robinson JT, Thorvaldsdóttir H, Winckler W,
53. Willer CJ, Li Y, Abecasis GR (2010) METAL: Guttman M, Lander ES, Getz G, Mesirov JP
fast and efficient meta-analysis of genomewide (2011) Integrative genomics viewer. Nat Bio-
association scans. Bioinformatics 26 technol 29(1):24–26
(17):2190–2191 61. Rivas MA, Pirinen M, Conrad DF, Lek M,
54. M€agi R, Morris AP (2010) GWAMA: software Tsang EK, Karczewski KJ, Maller JB, Kukurba
for genome-wide association meta-analysis. KR, DeLuca DS, Fromer M, Ferreira PG,
BMC Bioinformatics 11:288 Smith KS, Zhang R, Zhao F, Banks E,
55. Winkler TW, Kutalik Z, Gorski M, Lottaz C, Poplin R, Ruderfer DM, Purcell SM,
Kronenberg F, Heid IM (2015) EasyStrata: Tukiainen T, Minikel EV, Stenson PD, Cooper
evaluation and visualization of stratified DN, Huang KH, Sullivan TJ, Nedzel J; GTEx
genome-wide association meta-analysis data. Consortium; Geuvadis Consortium, Busta-
Bioinformatics 31(2):259–261 mante CD, Li JB, Daly MJ, Guigo R,
56. Huang Z, Lin H, Fellay J, Kutalik Z, Hubaux Donnelly P, Ardlie K, Sammeth M, Dermitza-
JP (2017) SQC: secure quality control for kis ET, McCarthy MI, Montgomery SB,
meta-analysis of genome-wide association Lappalainen T, MacArthur DG (2015)
studies. Bioinformatics 33(15):2273–2280 Human genomics. Effect of predicted
protein-truncating genetic variants on the
57. ENCODE Project Consortium (2012) An human transcriptome. Science 348
integrated encyclopedia of DNA elements in (6235):666–669
the human genome. Nature 489(7414):57–74
62. Machiela MJ, Chanock SJ (2015) LDlink: a
58. Roadmap Epigenomics Consortium, web-based application for exploring
Kundaje A, Meuleman W, Ernst J, Bilenky M, population-specific haplotype structure and
Yen A, Heravi-Moussavi A, Kheradpour P, linking correlated alleles of possible functional
Zhang Z, Wang J, Ziller MJ, Amin V, Whitaker variants. Bioinformatics 31(21):3555–3557
JW, Schultz MD, Ward LD, Sarkar A, Quon G,
Sandstrom RS, Eaton ML, Wu YC, Pfenning 63. Adzhubei IA, Schmidt S, Peshkin L, Ramensky
AR, Wang X, Claussnitzer M, Liu Y, Coarfa C, VE, Gerasimova A, Bork P, Kondrashov AS,
Harris RA, Shoresh N, Epstein CB, Sunyaev SR (2010) A method and server for
Gjoneska E, Leung D, Xie W, Hawkins RD,
predicting damaging missense mutations. Nat 77. Han B, Eskin E (2011) Random-effects model
Methods 7(4):248–249 aimed at discovering associations in meta-
64. Gamazon ER, Zhang W, Konkashbaev A, analysis of genome-wide association studies.
Duan S, Kistner EO, Nicolae DL, Dolan ME, Am J Hum Genet 88(5):586–598
Cox NJ (2010) SCAN: SNP and copy number 78. Han B, Eskin E (2012) Interpreting meta-
annotation. Bioinformatics 26(2):259–262 analyses of genome-wide association studies.
65. Munz M, Tönnies S, Balke WT, Simon E PLoS Genet 8(3):e1002555
(2015) Multidimensional gene search with 79. Hong J, Lunetta KL, Cupples LA, Dupuis J,
Genehopper. Nucleic Acids Res 43(W1): Liu CT (2016) Evaluation of a two-stage
W98–103 approach in trans-ethnic meta-analysis in
66. Drury TF, Horowitz AM, Ismail AI, Maertens genome-wide association studies. Genet Epi-
MP, Rozier RG, Selwitz RH (1999) Diagnos- demiol 40(4):284–292
ing and reporting early childhood caries for 80. M€agi R, Horikoshi M, Sofer T, Mahajan A,
research purposes. A report of a workshop Kitajima H, Franceschini N, McCarthy MI;
sponsored by the National Institute of Dental COGENT-Kidney Consortium, T2D-GENES
and Craniofacial Research, the Health Consortium, Morris AP (2017) Trans-ethnic
Resources and Services Administration, and meta-regression of genome-wide association
the Health Care Financing Administration. J studies accounting for ancestry increases
Public Health Dent 59(3):192–197 power for discovery and improves fine-
67. Clarkson J, O’Mullane D (1989) A modified mapping resolution. Hum Mol Genet 26
DDE Index for use in epidemiological studies (18):3639–3650
of enamel defects. J Dent Res 68(3):445–450 81. Maurano MT, Humbert R, Rynes E, Thurman
68. Westreich D (2012) Berkson’s bias, selection RE, Haugen E, Wang H, Reynolds AP,
bias, and missing data. Epidemiology 23 Sandstrom R, Qu H, Brody J, Shafer A,
(1):159–164 Neri F, Lee K, Kutyavin T, Stehling-Sun S,
69. Kraft P, Zeggini E, Ioannidis JP (2009) Repli- Johnson AK, Canfield TK, Giste E, Diegel M,
cation in genome-wide association studies. Stat Bates D, Hansen RS, Neph S, Sabo PJ,
Sci 24(4):561–573 Heimfeld S, Raubitschek A, Ziegler S,
Cotsapas C, Sotoodehnia N, Glass I, Sunyaev
70. Kraft P (2008) Curses—winner’s and other- SR, Kaul R, Stamatoyannopoulos JA (2012)
wise—in genetic epidemiology. Epidemiology Systematic localization of common disease-
19(5):649–651 associated variation in regulatory DNA. Sci-
71. Delaneau O, Marchini J, Zagury JF (2011) A ence 337(6099):1190–1195
linear complexity phasing method for 82. Shungin D, Cornelis MC, Divaris K,
thousands of genomes. Nat Methods 9 Holtfreter B, Shaffer JR, Yu YH, Barros SP,
(2):179–181 Beck JD, Biffar R, Boerwinkle EA, Crout RJ,
72. Delaneau O, Zagury JF, Marchini J (2013) Ganna A, Hallmans G, Hindy G, Hu FB,
Improved whole-chromosome phasing for dis- Kraft P, McNeil DW, Melander O, Moss KL,
ease and population genetic studies. Nat Meth- North KE, Orho-Melander M, Pedersen NL,
ods 10(1):5–6 Ridker PM, Rimm EB, Rose LM, Rukh G,
73. Das S, Forer L, Schönherr S, Sidore C, Locke Teumer A, Weyant RJ, Chasman DI,
AE, Kwong A, Vrieze SI, Chew EY, Levy S, Joshipura K, Kocher T, Magnusson PK, Mar-
McGue M, Schlessinger D, Stambolian D, Loh azita ML, Nilsson P, Offenbacher S, Davey
PR, Iacono WG, Swaroop A, Scott LJ, Cucca F, Smith G, Lundberg P, Palmer TM, Timpson
Kronenberg F, Boehnke M, Abecasis GR, NJ, Johansson I, Franks PW (2015) Using
Fuchsberger C (2016) Next-generation geno- genetics to test the causal relationship of total
type imputation service and methods. Nat adiposity and periodontitis: Mendelian ran-
Genet 48(10):1284–1287 domization analyses in the Gene-Lifestyle
74. Hancock DB, Levy JL, Gaddis NC, Bierut LJ, Interactions and Dental Endpoints (GLIDE)
Saccone NL, Page GP, Johnson EO (2012) Consortium. Int J Epidemiol 44(2):638–650
Assessment of genotype imputation perfor- 83. Shaffer JR, Feingold E, Wang X, Tcuenco KT,
mance using 1000 Genomes in African Ameri- Weeks DE, DeSensi RS, Polk DE, Wendell S,
can studies. PLoS One 7(11):e50610 Weyant RJ, Crout R, McNeil DW, Marazita
75. Stouffer SA (1949) Adjustment during army ML (2012) Heritable patterns of tooth decay
life. Princeton University Press, Princeton in the permanent dentition: principal compo-
76. Morris AP (2011) Transethnic meta-analysis of nents and factor analyses. BMC Oral Health
genomewide association studies. Genet Epide- 12:7
miol 35(8):809–822
84. Shaffer JR, Feingold E, Wang X, Weeks DE, 89. Zeng Z, Shaffer JR, Wang X, Feingold E,
Weyant RJ, Crout R, McNeil DW, Marazita Weeks DE, Lee M, Cuenco KT, Wendell SK,
ML (2013) Clustering tooth surfaces into bio- Weyant RJ, Crout R, McNeil DW, Marazita
logically informative caries outcomes. J Dent ML (2013) Genome-wide association studies
Res 92(1):32–37 of pit-and-fissure- and smooth-surface caries in
85. Morelli T, Moss KL, Beck J, Preisser JS, Wu D, permanent dentition. J Dent Res 92
Divaris K, Offenbacher S (2017) Derivation (5):432–437
and validation of the periodontal and tooth 90. Hu YJ, Li Y, Auer PL, Lin DY (2015) Integra-
profile classification system for patient stratifi- tive analysis of sequencing and array genotype
cation. J Periodontol 88(2):153–165 data for discovering disease associations with
86. Recent advances in oral health. Report of a rare mutations. Proc Natl Acad Sci U S A 112
WHO Expert Committee (1992) World (4):1019–24
Health Organ Tech Rep Ser 826:1–37 91. Divaris K, North KE, Slade GD, Barros SP,
87. Zeng Z, Feingold E, Wang X, Weeks DE, Moss K, Beck JD, Offenbacher S (2014)
Lee M, Cuenco DT, Broffitt B, Weyant RJ, Genome-wide association study of tooth mor-
Crout R, McNeil DW, Levy SM, Marazita bidity. J Dent Res 92(Spec Issue B):190702
ML, Shaffer JR (2014) Genome-wide associa- 92. Eke PI, Page RC, Wei L, Thornton-Evans G,
tion study of primary dentition pit-and-fissure Genco RJ (2012) Update of the case defini-
and smooth surface caries. Caries Res 48 tions for population-based surveillance of peri-
(4):330–338 odontitis. J Periodontol 83(12):1449–1454
88. Shaffer JR, Feingold E, Wang X, Lee M, 93. Schaefer AS, Richter GM, Groessner-
Tcuenco K, Weeks DE, Weyant RJ, Crout R, Schreiber B, Noack B, Nothnagel M, El Mokh-
McNeil DW, Marazita ML (2013) GWAS of tari NE, Loos BG, Jepsen S, Schreiber S (2009)
dental caries patterns in the permanent denti- Identification of a shared genetic susceptibility
tion. J Dent Res 92(1):38–44 locus for coronary heart disease and periodon-
titis. PLoS Genet 5(2):e1000378
Chapter 39
Measurement of Early Childhood Oral Health for Research

Purposes: Dental Caries Experience and Developmental
Defects of the Enamel in the Primary Dentition
Jeannie Ginnis, Andrea G. Ferreira Zandoná, Gary D. Slade, John Cantrell,
Mikafui E. Antonio, Bhavna T. Pahel, Beau D. Meyer, Poojan Shrestha,
Miguel A. Simancas-Pallares, Ashwini R. Joshi, and Kimon Divaris
Abstract
Epidemiological investigations of early childhood oral health rely upon the collection of high-quality
clinical measures of health and disease. However, ascertainment of valid and accurate clinical measures
presents unique challenges among young, preschool-age children. The paper presents a clinical research
protocol for the conduct of oral epidemiological examinations among children, implemented in ZOE 2.0, a
large-scale population-based genetic epidemiologic study of early childhood caries (ECC). The protocol has
been developed for the collection of information on tooth surface-level dental caries experience and tooth-
level developmental defects of the enamel in the primary dentition. Dental caries experience is recorded
using visual criteria modified from the International Caries Detection and Assessment System (ICDAS),
and measurement of developmental defects is based upon the modified Clarkson and O’Mullane Develop-
mental Defects of the Enamel Index. After a dental prophylaxis (toothbrushing among all children and
flossing as needed), children’s teeth are examined by trained and calibrated examiners in community
locations, using portable dental equipment, compressed air, and uniform artificial light and magnification
conditions. Data are entered directly onto a computer using a custom Microsoft Access-based data entry
application. The ZOE 2.0 clinical protocol has been implemented successfully for the conduct of over 6000
research examinations to date, contributing phenotype data to downstream genomics and other “omics”
studies of ECC and DDE, as well as traditional clinical and epidemiologic dental research.
Key words Dental caries, Children, Clinical research, Protocol, Caries diagnosis, Early childhood
caries, Developmental defects of the enamel
1 Introduction
Epidemiological investigation of early childhood oral health relies

upon the collection of high-quality clinical measures of health and
disease. This, in turn, requires a clinical research protocol to ensure
robust and unbiased experimental design, methodology, analysis,
interpretation, and reporting of results. In the research studies of
511
512 Jeannie Ginnis et al.
oral health domain, ascertainment of valid and accurate clinical

measures presents unique challenges among young, preschool-age
children. This is due to the reduced ability (e.g., limited attention
span, acute anxiety, etc.) of very young children to cope with and
cooperate with clinical research procedures compared to older
children and adults. Nevertheless, clinical dental research is not
only feasible in this young age group but also necessary to answer
important biological, clinical, and public health research
questions [1].
Understanding the entire spectrum of social, environmental,
behavioral, and biological determinants of early childhood caries
(ECC) [2] remains a priority. ECC is an early-onset, severe form of
dental caries that confers substantial impacts to affected children,
their families, the health system, and the society in general. It is a
persistent clinical and dental public health problem that is known to
be influenced by social determinants of health and mediated by oral
hygiene, fluoride exposure, and diet [3] but also has a substantial
heritable component [4]. Developmental defects of the enamel
(DDE) are non-carious enamel conditions, comprising a heteroge-
neous group of defects of various severities and diverse etiologies
(i.e., hypoplastic defects of the enamel versus fluorosis). Impor-
tantly, DDE occur during dental development (i.e., prior to tooth
eruption and exposure to protective and harmful exposures), and
because they can increase one’s susceptibility to dental caries [5, 6],
their study within the context of ECC heritability can be
informative.
Genetic studies of dental caries in the primary dentition caries
have had mixed results [7, 8], and a knowledge gap exists on the
genomic basis of ECC [9]. To address this knowledge gap and add
to the evidence base of social, behavioral, and biological determi-
nants of ECC, the investigators are undertaking the ZOE-G4S
(Zero-Out Early Childhood Caries “Genes for Smiles”)—an
NIH-funded genetic epidemiologic study conducted among a
large population-based sample of preschool-age children enrolled
in Head Start centers in North Carolina. The paper presents the
study’s clinical protocol for the conduct of research examinations
among children in the primary dentition (e.g., ages 2–5), designed
to detect and record surface-level dental caries experience and
tooth-level developmental defects of the enamel. The phenotype
data collected using this protocol are then carried forward to geno-
mics [15] and microbiome [16] studies, as well as other traditional,
clinical, and epidemiological dental investigations.
Clinical examiners (registered dental hygienists or dentists) are
trained and calibrated for the dental caries examinations using a
sequence of online International Caries Detection and Assessment
System (ICDAS) modules, clinical photos, natural extracted teeth,
and clinical examinations. Tooth surface caries diagnoses are based
on ICDAS criteria [10, 11] made at the levels of health (ICDAS
Measurement of Dental Caries and Developmental Defects of the Enamel 513
code, 0), early stage (ICDAS codes, 1–2), and established/severe

stage (ICDAS codes, 3–6). A dmfs (decayed, missing due to caries
and filled/restored primary tooth surfaces, ranging between 0 and
88) index is created as a sum of surface-level conditions, with
dmfs > 0 indicating an ECC case [2], which is the primary study
outcome. The presence of sealants and intracoronal and full-
coverage restorations is also recorded at the surface level. Inter-
examiner reliability in caries classification is determined by compar-
ing uncalibrated examiners with a reference golden standard exam-
iner. The calibration includes examination of ~25 children by all
examiners, including uncalibrated ones and the gold standard, and
re-examination of ~8. Examiner performance is evaluated by the
weighted kappa statistic, indicating strength of agreement with the
gold standard examiner. Achieving a weighted kappa of at least
0.65, a rigorous level of inter-examiner reliability, is required for
all examiners following their initial training and calibration and at
annual re-evaluation sessions. The corresponding threshold for
intra-examiner reliability is weighted kappa at least 0.75 [12].
Developmental defects of the enamel are measured at the tooth
level using Clarkson and O’Mullane’s modified DDE epidemio-
logic index [13]. Several other clinical measures are obtained as part
of the research examination, including anthropometry (i.e., height,
weight, Body Mass Index - BMI, BMI percentile for age and sex),
extraoral characteristics (i.e., profile, lip competence), intraoral
occlusal characteristics (i.e., molar/canine classification, overjet,
overbite), dental trauma (using a modified Ellis index [14]), and a
summary of treatment needs. With supervision and/or assistance
from research staff, children’s teeth are cleaned using a toothbrush
with no toothpaste; and flossing when needed. The aim is to
improve measurement and, in the process, provide instruction in
dental hygiene while helping desensitize children who might be
fearful of dental procedures. Before tooth cleaning, saliva and
supragingival plaque samples are collected to enable downstream
studies involving the human genome and oral microbiome, accord-
ing to procedures detailed in [15, 16].
2 Materials
2.1 Clinical 1. Foldable dental chair (Aseptico® ADC-01P-RED, Aseptico

Examination Inc., Woodinville, WA).
Equipment and 2. Foldable hydraulic dentist stool (Aseptico® ADC-10, Aseptico
Instruments Inc., Woodinville, WA).
3. Dental instrument tray (Aseptico® ATC-03CF, Aseptico Inc.,
Woodinville, WA).
4. Air compressor (Fig. 1; custom modified TC-20 compressor,
TCP Global, San Diego, CA) with air syringe and
disposable tips.
Fig. 1 Custom modified air compressor
5. Examiner magnifying loupes with headlight (Orascoptic

XV1™ Loupe + Light, Orascoptic, Middleton, WI).
6. Pediatric toothbrushes.
7. Flossers.
8. Disposable dental examination mirrors.
9. US/CPTIN round-ended plastic periodontal probes with
markings at 2/3/4/5/7/9 mm.
2.2 Clinical 1. Disposable gowns.

Examination Supplies 2. Gloves.
3. Masks.
4. Bibs.
5. Handheld mirrors.
6. Disposable plastic tips and sleeves/covers for air syringe.
7. Barrier tape.
8. Plastic chair and tray covers.
9. Surface disinfecting wipes.
10. Hand sanitizer.
11. 2 2 gauze.
12. Paper towels.
13. Sterile pouches for used probes.
2.3 Anthropometric 1. Portable digital stadiometer (Seca® 213 or Seca® 264, Seca
Measurement GmbH & Co. KG, Hamburg Germany).
Equipment and 2. Portable digital scale (DS6150, Doran® Remote Indicator
Supplies Scale, Doran, Batavia, IL).
3. Carrying case for portable digital stadiometer and scale (Seca®

414, Seca GmbH & Co. KG, Hamburg Germany).
4. iPad (Apple Inc., Cupertino, CA, USA) tablet with “Child-
BMI” app (Tactio Health Group) for pediatric BMI percentile
calculation.
5. Printed “subject cards” for intermediate recording of height,
weight, BMI, and BMI percentile values prior to electronic
data capture.
2.4 Electronic Data 1. Laptop.

Recording and Other 2. Microsoft Access-based (Microsoft Corp., Redmond, WA,
Study Source USA), custom-written data entry application (DEA).
Documents
3. Handheld barcode reader (for scanning of barcode labels with
participant IDs).
4. Printed screenshots of DEA data entry screen (as backup).
5. External monitor for examiner.
6. Power strips, extension cords, and adapters.
7. Blue masking (floor) tape.
8. Encrypted, portable USB storage device (e.g., IronKey™,
Kingston Technology Company, Inc., Fountain Valley, CA).
2.5 Biospecimen 1. Saliva collection for DNA extraction is detailed in [15].

Collection 2. Supragingival plaque collection for microbiome analyses (i.e.,
metagenomics, metatranscriptomics, and metabolomics) is
detailed in [16].
3 Methods
Research examination teams use the mobile dental equipment,

instruments, and supplies to set up clinical examination stations
(Fig. 2) where there is suitable space and a power outlet in commu-
nity locations (e.g., classrooms). Each examiner has custom-fit
magnifying loupes with headlights at their disposal (Subheading
2.1) and is trained to use basic child behavior guidance and man-
agement techniques routinely used in pediatric dentistry, including
tell-show-do (TSD), positive reinforcement, negative reinforce-
ment, voice control, modeling, and labeling. Children whose legal
guardians have provided informed consent are examined—at least
30, preferably 60 min—after they have had breakfast or snack. The
families (and teachers, where applicable) are instructed not to brush
the children’s teeth the morning before the examination—the
examiner cleans the teeth with a toothbrush (no toothpaste) and
floss (when needed) after biospecimen collection, and the teeth are
dried before the examination. The examiner completes the exam,
Fig. 2 Example of a clinical research setup in a nonclinical setting
and all data are directly recorded electronically by a designated

recorder using the DEA. The procedures are carried out as follows:
3.1 Preclinical 1. Greet the child addressing them by their name and ideally
Procedures and kneeling down to their eye level.
Assessments 2. Explain the procedures that will follow in an age-appropriate
manner.
3. Offer stickers at each examination “station” beginning at initial
contact.
4. Record height and weight, and compute BMI and BMI per-
centile for age and sex.
5. Collect saliva sample with OG-575 kit (DNA Genotek.,
Ottawa, Ontario, Canada). This can be accomplished by the
child sitting upright in any (regular or dental chair).
6. Collect supragingival plaque samples using the sterile tooth-
picks in a semi-reclined position, ideally a dental chair. One
sample (plaque sample A) from the upper right dentition (facial
surfaces of #A, #B, #C, #D, and #E according to the universal
nomenclature system or #55, #54, #53, #52, and #51 accord-
ing to the FDI system) and another sample (plaque sample B)
from the upper left dentition (facial surfaces of #F, #G, #H, #I,
and #J or #61, #61, #63, #64, and #61).
7. Store plaque A in an RNAlater tube and plaque B in a Cryovial
to freeze immediately, until transferred to a lab and stored at
80 C.
3.2 Clinical 1. Clean teeth with a toothbrush, and provide oral hygiene
Procedures and instruction for all children. Clean interproximal areas with
Assessments of Dental dental floss as needed.
Caries and DDE 2. Record extraoral characteristics as follows: profile (1 ¼ straight,
2 ¼ convex, 3 ¼ concave, 9 ¼ unable to assess) and lip compe-
tence (1 ¼ competent, <3 mm distance between upper and
lower lips and no mentalis muscle strain; 2 ¼ incompetent,
3 mm distance between upper and lower lips or mentalis
muscle strain; 9 ¼ unable to assess).
3. Record intraoral characteristics as follows: overjet (mm;
999 ¼ unable to assess), overbite (percent coverage of lower
incisors from the upper incisors; 999 ¼ unable to assess),
right/left molar anterior-posterior classification (1 ¼ flush,
2 ¼ mesial step, 3 ¼ distal step, 9 ¼ unable to assess), and
right/left canine anterior-posterior classification (1 ¼ class I,
2 ¼ class II, 3 ¼ class III, 9 ¼ unable to assess).
4. Dry each tooth to be examined for surface-level dental caries
lesions or other conditions with compressed air
(or alternatively with a 2 2 gauze), and visualize with artificial
light, magnification, and disposable mirror. Recording is based
on a two-step system: first provide a tooth-level code (Table 1)
and then, as indicated, four or five surface-level codes
Table 1
Tooth-level diagnostic and classification codes for recording of dental
caries
Code Explanation
1 Sound, no decay, or restorationsa
2 Sealedb
3 Decayed or filled (restored)b
4 Crown (stainless steel or other)a
5 Missing due to cariesa
6 Exfoliateda
7 Missing due to traumaa
8 Permanent tooth presenta
9 Unable to scorea
a
For the other tooth-level codes, tooth surface codes are auto-filled by the DEA, as
follows: tooth-level code 1, sound (auto-filled surface-level codes, 0); tooth-level code
4, crown (auto-filled surface-level code, 44); tooth-level code 5, missing due to caries
(auto-filled surface-level code, 55); tooth-level codes 6, 7, 8, and 9, exfoliated, missing
due to trauma, permanent tooth present, and unable to score (auto-filled surface-level
code, 99)
b
Tooth surface-level codes are entered by the examiner only for tooth-level codes
2 and 3. See Table 2
Table 2
Surface-level diagnostic and classification codes for recording of dental caries
Code Explanation
0 Sound, no caries lesion (ICDAS, 0)
50 Restored, no caries lesion
80 Sealed, no caries lesiona
Caries
10 Arrested enamel lesion (enamel-only)b
11 Early stage caries lesion (ICDAS, 1–2)
12 Established/severe caries lesion (ICDAS, 3–6)
Restored and separate caries lesion
30 Restored and separate arrested enamel lesion (enamel-only)b
31 Restored and separate early-stage caries lesion (ICDAS, 1–2)
32 Restored and separate established/severe caries lesion (ICDAS, 3–6)
Restored and recurrent caries lesion
40 Restored and recurrent arrested enamel lesion (enamel-only)b
41 Restored and recurrent early-stage caries lesion (ICDAS, 1–2)
42 Restored and recurrent established/severe caries lesion (ICDAS, 3–6)
Sealed and separate caries lesion
60 Sealed and separate arrested enamel lesion (enamel-only)b
61 Sealed and separate early-stage caries lesion (ICDAS, 1–2)
62 Sealed and separate established/severe caries lesion (ICDAS, 3–6)
Sealed and recurrent caries lesion
70 Sealed and recurrent arrested enamel lesion (enamel-only)b
71 Sealed and recurrent early-stage caries lesion (ICDAS, 1–2)
72 Sealed and recurrent established/severe caries lesion (ICDAS, 3–6)
Auto-filled codes
44 Surface of tooth with a crown, i.e., if tooth code ¼ 4
55 Surface of tooth that has been extracted due to caries, i.e., if tooth code ¼ 5
99 Surface of tooth that has exfoliated, is missing due to trauma, a permanent tooth is present, or is
unable to score for any other reason, i.e., if tooth code ¼ 6, 7, 8, or 9
a
Any evidence of sealant material present (i.e., partially retained) is recorded as sealant
b
Caries lesion activity is only considered for enamel-level non-cavitated (i.e., white spot) lesions
(Table 2). The purpose of this system is to (a) expedite data

collection in cases where all surfaces within a tooth receive the
same diagnosis (i.e., healthy, crowned, extracted due to caries,
missing due to trauma, permanent tooth is present, unable to
assess) and (b) allow the implementation of logical checks, thus
minimizing data entry errors. Five surfaces (occlusal, facial, lin-
gual, mesial, and distal) are assessed for molar teeth and four
surfaces (facial, lingual, mesial, and distal) for anterior teeth.
5. Examination for DDE is best done on non-desiccated teeth,
and for this reason, the child is asked to close their mouth and
swallow or lick their teeth after the dental caries examination.
Examine only the facial/buccal surface of all teeth for
Table 3
Diagnostic and classification codes for recording of type and extent of
developmental defects of the enamel
Code Explanation
Type
0 Normal (no DDE)
1 Demarcated opacity
2 Diffused opacity
3 Hypoplasia
4 Demarcated opacity + diffused opacity
5 Demarcated opacity + hypoplasia
6 Diffused opacity + hypoplasia
7 Demarcated opacity + diffused opacity + hypoplasia
8 Other defect
9 Unable to score
Extent
0 Normal (no DDE)
1 < 1/3 of facial/buccal tooth surface
2 At least 1/3 but less than 2/3 of facial/buccal tooth surface
3 At least 2/3 of facial/buccal tooth surface
9 Unable to score
non-carious enamel defects that are of 1 mm diameter, using

the type and extent codes listed in Table 3. If more than one
condition is present (e.g., diffuse opacity and hypoplasia),
record the extent of the largest one. If more than two-thirds
of the tooth’s facial/buccal surface is heavily restored, badly
decayed, or fractured, then do not assess it for DDE (i.e.,
9 ¼ unable to assess).
3.3 Measurement of 1. Examine the four primary maxillary incisors for evidence of
Other Clinical dental trauma using the modified Ellis [14] classification cri-
Characteristics: teria listed in Table 4.
Trauma, Treatment 2. Provide a behavior assessment score at the end of the examina-
Needs, Behavior, tion to provide a global measure of the child’s cooperation
Completion, and using the Frankl behavior scale [17], as follows: 4 ¼ completely
Examination Notes cooperative, enjoys the process; 3 ¼ cooperative, somewhat
reluctant/shy; 2 ¼ uncooperative, very reluctant/shy; and
1 ¼ completely uncooperative. In cases where part or the
entirety of the research examination was not done due to
noncooperation, a score of 1 is given.
3. Provide a “treatment needs” score, according to the level and
urgency of dental treatment that the child requires, as follows:
1 ¼ no obvious problems were found today, establish or main-
tain a dental home and continue routine dental visits; 2 ¼ possi-
ble problems were found today, please check at the next dental
visit; and 3 ¼ problems were found today that require treat-
ment as soon as possible.
Table 4
Diagnostic and classification codes for recording of dental trauma
Code Explanation
0 No evidence of trauma
1 Simple crown fracture, involving little or no dentin
2 Extensive crown fracture, involving considerable dentin surface, without pulp exposure
3 Extensive crown fracture, involving considerable dentin surface, with pulp exposure
4 Displacement of the tooth due to trauma
5 Necrotic/discolored tooth (due to trauma) w/wo crown fracture
6 Total tooth loss due to trauma
7 Unable to assess
4. Provide a “completion” code indicating whether all planned

research procedures were completed at this visit: 1 ¼ completed
and 2 ¼ not complete. A built-in check routine in the DEA will
prevent entering “1” if a data field is incomplete.
5. Provide free-text comments in the appropriate comment box, if
and as needed. In this section, general remarks protocol devia-
tions (e.g., changes in procedures due to behavior/coopera-
tion) and any extraordinary findings can be recorded. A
comment is mandated if a partial exam (i.e., “not complete”
code) was conducted.
4 Notes
1. Explaining all procedures to the participating children using the

“tell-show-do” technique and actively involving them in the
process, where they can even hold a hand mirror to watch and,
for example, “help the examiner count teeth,” are very helpful
strategies—especially in cases of reluctant or anxious children.
2. In cases of uncertainty about the diagnosis/classification of a
condition, prefer “under-calling” it. For instance, if uncertain
whether an early-stage (incipient) caries lesion is present or not,
prefer to record the lesser diagnosis, which—in this case—is
health (e.g., surface code, 0). Apart from resulting in a more
conservative estimation of disease burden, this rule is helpful in
training and calibration, so that examiners have a uniform rule
for resolving uncertainties.
3. Multiple traits besides the ECC case status (binary) and the
dmfs index (0–88) can be constructed using the diagnostic
codes collected as part of the protocol. The construction of
six example traits is presented in Table 5, where trait 3 is the
Table 5
Examples of six person-level caries traits (disease indices and definitions) that can be constructed using the surface-level caries diagnosis codes used in
this protocol
Trait 1 Trait 2 Trait 3 Trait 4 Trait 5 Trait 6 Trait 7a Trait 8a Trait 9a Trait 10a
Caries
Two- Established/ Caries experience, “Healthy,”
Composite level ECC severe experience, established/ “Restored” “Unrestored” “Healthy,” non- Surface
index disease definition disease two-level severe disease disease sealed sealed codesb Description
0 0 0 0 0 0 0 0 0 1 0 Sound
1 0 0 0 0 0 0 0 1 0 80 Sealant
2 1 1 0 1 0 0 0 0 1 10, 11 Caries: early stage
60, 61 (arrested or active)
70, 71
2 1 1 0 2 1 1 0 0 0 30c, 31c Recurrent or Secondary
40c, 41c Caries: early stage
(arrested or active)
3 2 1 1 2 1 0 1 0 0 12, 32c, Caries: established/severe
42c,
62, 72
4 2 1 1 2 1 1 0 0 0 50 Restored
5 2 1 1 2 1 1 0 0 0 44 Crowned
[auto-
filled]
6 2 1 1 2 1 1 0 0 0 55 Missing due to caries
[auto-
filled]
Additional person-level indicator variable definitions: restorativetx = 1 if one or more surfaces are in (30, 31, 32, 40, 41, 42, 50, 44, 55), else restorativetx = 0; sealantstx = 1 if one or
more surfaces are in (80, 60, 61, 62, 70, 71, 72), else sealantstx = 0
a
For Traits 7–10 disease is considered at the established/severe level
Measurement of Dental Caries and Developmental Defects of the Enamel
b
Surfaces with code: 99 (when tooth code is 6, 7, 8, or 9) are excluded
c
These surfaces have been restored, and thus can be grouped with trait 4 if the interest is examining “disease severity”; if the interest is in active disease, then these surfaces can be
counted according to the lesion classification, i.e., as early stage (Trait 2) or established/severe (Trait 3)
521
“classic” ECC definition and trait 4 is the same case status

definition not considering early-stage lesions.
4. The custom modified air compressor referred to in Subheading
2.1 has proven to be an excellent solution for the provision of
compressed air during the visual dental caries examinations in
the field—it is lightweight and silent and produces sufficient
airflow.
5. Consider a tooth present if any part of it has penetrated the
mucosa, and record any condition that can be detected on its
erupted portion.
6. Overjet measurements can be performed if any pair of upper-
lower primary central incisors is present (e.g., #F-O or #E-P).
7. Placing multiple layers of plastic covering on dental chair and
tray saves time when turning over the setup for the next
participant.
Acknowledgment
This work was supported by a grant from the National Institutes of

U01-DE025046.
References
1. Casamassimo PS, Lee JY, Marazita ML, 5. Seow WK (2014) Developmental defects of
Milgrom P, Chi DL, Divaris K (2014) Improv- enamel and dentine: challenges for basic sci-
ing children’s oral health: an interdisciplinary ence research and clinical management. Aust
research framework. J Dent Res 93 Dent J 59(Suppl 1):143–154
(10):938–942 6. Caufield PW, Li Y, Bromage TG (2012 Jun)
2. Drury TF, Horowitz AM, Ismail AI, Maertens Hypoplasia-associated severe early childhood
MP, Rozier RG, Selwitz RH (1999 Summer) caries—a proposed definition. J Dent Res 91
Diagnosing and reporting early childhood car- (6):544–550
ies for research purposes. A report of a work- 7. Vieira AR, Modesto A, Marazita M (2014)
shop sponsored by the National Institute of Caries: review of human genetic research. Car-
Dental and Craniofacial Research, the Health ies Res 48:491–506
Resources and Services Administration, and 8. Shaffer JR, Wang X, Feingold E et al (2011)
the Health Care Financing Administration. J Genome-wide association scan for childhood
Public Health Dent 59(3):192–197 caries implicates novel genes. J Dent Res 90
3. Divaris K (2016) Predicting dental caries out- (12):1457–1462
comes in children: a “risky” concept. J Dent 9. Divaris K (2017) Precision dentistry in early
Res 95(3):248–254 childhood: the central role of genomics. Dent
4. Ballantine J, Carlson JC, Zandona A, Agler CS, Clin North Am 61(3):619–625
Zeldin LP, Rozier G, Roberts MW, Basta PV, 10. Pitts N (2004) “ICDAS”—an international
Luo J, Antonio-Obese ME, McNeil D, system for caries detection and assessment
Weyant R, Crout RJ, Slayton R, Levy S, Shaffer being developed to facilitate caries epidemiol-
JR, Marazita ML, North KE, Divaris K (2018) ogy, research and appropriate clinical manage-
Exploring the genomic basis of early childhood ment. Community Dent Health 21
caries: a pilot study. Int J Paediatr Dent 28 (3):193–198
(2):217–225
11. Ismail AI, Sohn W, Tellez M, Amaya A, Sen A, 15. Agler AS, Shungin D, Ferreira Zandoná AG,
Hasson H, Pitts NB (2007) The International Basta PV, Luo J, Cantrell J, Pahel Jr. TD,
Caries Detection and Assessment System Meyer BD, Shaffer JR, Sch€aefer AS, North
(ICDAS): an integrated system for measuring KE, Divaris K (this volume) Protocols, meth-
dental caries. Community Dent Oral Epide- ods and tools for genome-wide association
miol 35(3):170–178 studies (GWAS) of dental traits. Methods Mol
12. Banting DW, Amaechi BT, Bader JD, Biol 1922:493–509
Blanchard P, Gilbert GH, Gullion CM, Hol- 16. Shungin D, Basta PV, Azcarate-Peril AM, Fer-
land JC, Makhija SK, Papas A, Ritter AV, Singh reira Zandoná AG, Cho H, Wu D, Ginnis J,
ML, Vollmer WM (2011) Examiner training Cantrell J, Pahel Jr. TD, Roach J, Divaris K
and reliability in two randomized clinical trials (this volume) Protocol for clinical, laboratory
of adult dental caries. J Public Health Dent 71 and bioinformatics pipelines for oral micro-
(4):335–344 biome studies of dental caries involving meta-
13. Clarkson J, O’Mullane D (1989) A modified genomics, metatrascriptomics and
DDE Index for use in epidemiological studies metabolomics. Methods Mol Biol 1922:
of enamel defects. J Dent Res 68(3):445–450 525–548
14. Ellis RG, Davey KW (1970) The classification 17. Frankl SN, Shiere FR, Fogels HR (1962)
and treatment of injuries to the teeth of chil- Should the parent remain with the child in the
dren: a reference manual for the dental student dental operatory? J Dent Child 29:150–163
and the general practitioner. Year Book Medi-
cal Publishers, 5th Edition, Chicago
Chapter 40
The Supragingival Biofilm in Early Childhood Caries: Clinical

and Laboratory Protocols and Bioinformatics Pipelines
Supporting Metagenomics, Metatranscriptomics, and
Metabolomics Studies of the Oral Microbiome
Kimon Divaris, Dmitry Shungin, Adaris Rodrı́guez-Cortés,
Patricia V. Basta, Jeff Roach, Hunyong Cho, Di Wu,
Andrea G. Ferreira Zandoná, Jeannie Ginnis, Sivapriya Ramamoorthy,
Jason M. Kinchen, Jakub Kwintkiewicz, Natasha Butz, Apoena A. Ribeiro,
and M. Andrea Azcarate-Peril
Abstract
Early childhood caries (ECC) is a biofilm-mediated disease. Social, environmental, and behavioral deter-
minants as well as innate susceptibility are major influences on its incidence; however, from a pathogenetic
standpoint, the disease is defined and driven by oral dysbiosis. In other words, the disease occurs when the
natural equilibrium between the host and its oral microbiome shifts toward states that promote deminerali-
zation at the biofilm-tooth surface interface. Thus, a comprehensive understanding of dental caries as a
disease requires the characterization of both the composition and the function or metabolic activity of the
supragingival biofilm according to well-defined clinical statuses. However, taxonomic and functional
information of the supragingival biofilm is rarely available in clinical cohorts, and its collection presents
unique challenges among very young children. This paper presents a protocol and pipelines available for the
conduct of supragingival biofilm microbiome studies among children in the primary dentition, that has
been designed in the context of a large-scale population-based genetic epidemiologic study of ECC. The
protocol is being developed for the collection of two supragingival biofilm samples from the maxillary
primary dentition, enabling downstream taxonomic (e.g., metagenomics) and functional (e.g., transcrip-
tomics and metabolomics) analyses. The protocol is being implemented in the assembly of a pediatric
precision medicine cohort comprising over 6000 participants to date, contributing social, environmental,
behavioral, clinical, and biological data informing ECC and other oral health outcomes.
Key words Dental caries, Microbiome, Transcriptome, Metabolome, Children, Protocol
Electronic supplementary material: The online version of this chapter (https://doi.org/10.1007/978-1-4939-

9012-2_40) contains supplementary material, which is available to authorized users.
525
526 Kimon Divaris et al.
1 Introduction
Early childhood caries (ECC), as all other taxonomic entities of

dental caries, is a biofilm-mediated, dysbiotic disease [1, 2]. Social,
environmental, and behavioral determinants, as well as innate sus-
ceptibility, are known major influences on its incidence [2, 3];
however, from a pathogenetic standpoint, the disease is defined
and driven by oral dysbiosis. In other words, the disease occurs
when the natural equilibrium between the host and its oral micro-
biome shifts toward states that promote demineralization at the
biofilm-tooth surface interface [4–6]. Conceptualizing ECC as an
oral-ecological imbalance is a shift away from currently used clinical
definitions [7]. The latter are undoubtedly useful for classifying and
reporting clinical outcomes of the disease process in clinical prac-
tice, research, as well as dental public health and surveillance.
However, for the realization of precision oral health care [2, 8],
the biological processes underlying these clinical manifestations
must be understood and defined.
Measures of the composition and activity of the supragingival
biofilm can be regarded as endophenotypes [8], similar to that of
cholesterol and systemic inflammatory markers in the context of
cardiovascular health-proximal, sensitive traits of pathogenetic
activity. A comprehensive understanding of ECC—and all presen-
tations of dental caries—as a disease process (versus its clinical
manifestation) requires the characterization of both the composi-
tion and the function or metabolic activity of the supragingival
biofilm according to well-defined clinical statuses. However, taxo-
nomic and functional information of the supragingival biofilm is
rarely available in clinical cohorts, and its collection presents unique
challenges among very young children. This paper presents a pro-
tocol and pipelines available for supragingival biofilm microbiome
studies among children in the primary dentition, implemented in a
large-scale population-based genetic epidemiologic study of early
childhood caries (ECC). The protocol has been developed for the
collection of two supragingival biofilm (plaque) samples from the
maxillary primary dentition, enabling downstream taxonomic (e.g.,
metagenomics [9]) and functional (e.g., transcriptomics [10] and
metabolomics [11]) analyses. We are implementing this protocol in
the assembly of a pediatric precision medicine cohort currently
comprising over 6000 participants, contributing a multitude of
data informing ECC and other oral health outcomes. The project
is the ZOE-G4S (Zero-Out Early Childhood Caries-“Genes for
Smiles”; ZOE 2.0) study—a NIH-funded genetic epidemiologic
investigation among a large population-based sample of preschool-
age children enrolled in Head Start centers in North Carolina. A
key feature of the protocol is that it outlines biofilm collection and
Biofilm Studies in ECC: Microbiome, Transcriptome and Metabolome 527
storage procedures that can be done under field conditions, hours

or days away from a clinic or laboratory.
Supragingival biofilm is being collected from participating 3–5-
year-old children by clinical examiners (registered dental hygienists
or dentists) who are trained and calibrated for dental caries experi-
ence assessment. Study staff instruct families and school teachers
not to brush the children’s teeth the morning of the study visit.
After saliva (for genomic DNA extraction purposes) and biofilm
collection, the examiners brush and floss children’s teeth, and,
besides dental caries experience, they record developmental defects
of the enamel, anthropometric measures, extra-oral characteristics,
dental occlusion classification, dental trauma prevalence, and a
summary of treatment needs.
2 Materials, Equipment, and Software
The biofilm sample collection is done in the context of a main

clinical protocol that entails measurement of dental caries experi-
ence and other clinical indices, requiring a full dental setup, as
detailed in Subheading 2.1. Conceivably, dental biofilm collection
can be done without a full dental setup (e.g., with the child sitting
on a regular versus a dental chair)—however, in most instances,
biofilm collection will be accompanied by measurement of clinical
characteristics, and, for this reason, a dental setup will be necessary
and available. The samples are frozen at the collection site and
transferred to a core lab facility for long-term storage or nucleic
acid (NA) extraction.
Several NA extraction methods from biofilms are available; our
team has been testing and optimizing both manual and high-
throughput (96-well plate) purification methods. After purification
and quantitation, NA preps are transferred to a sequencing facility
and carried forward to whole genome shotgun (WGS) and RNA
sequencing. Our group has verified and reported that the described
procedures lead to good quality and informative transcriptome
(unpublished data) and metagenomics results [12] with respect to
oral health and ECC statuses. Bioinformatics pipelines are then
executed on a research computing cluster to carry out steps
described in Subheading 3.6 as well as additional data management
and statistical analyses.
Biofilm samples collected for metabolomics analyses are cur-
rently bio-banked in 80 C, while range-finding and pilot studies
are under way in collaboration with Metabolon®. Metabolon uses
multiple mass spectrometry methods, a large reference library of
authenticated metabolite standards and a suite of patented infor-
matics and quality control software in the DiscoveryHD4™ plat-
form [13–17]. A pilot study including metabolomics analysis of
300 ZOE 2.0 biofilm samples identified 503 compounds of known
identity (named biochemicals) in these samples.
2.1 Clinical l Foldable dental chair (Aseptico® ADC-01P-RED).

Examination l Foldable hydraulic dentist stool (Aseptico® ADC-10).
Equipment and l Dental instrument tray (Aseptico® ATC-03CF).
Instruments
l Examiner magnifying loupes with headlight (Orascoptic XV1™
Loupe + Light).
2.2 Clinical l Disposable gowns.

Examination Supplies l Gloves.
l Masks.
l Bibs.
l Disposable dental examination mirrors.
l Barrier tape.
l Plastic chair and tray covers.
l Surface disinfecting wipes.
l Hand sanitizer.
2.3 Supragingival l CoolBox™ XT (BioCision®; BCS-575) or 2XT (BioCision®;

Biofilm Collection BCS-573) workstations, with Freezing Cores (BioCision®;
Equipment BCS-512), and M24 (BioCision®; BCS-535) and CFT24 (Bio-
Cision®; BCS-534) CoolRack® for on-site immediate freezing
and transportation of biofilm samples.
l Fridge freezer 50 qt. (ARB®; item #10800472) for long-
distance transport of samples.
l Power cord (ARB®; item #10910076) for in-car power supply of
portable freezer.
l Canvas travel bag (ARB®; item #10900013), optional for porta-
ble freezer.
l Fridge remote monitor (ARB®; item #10900026), optional for
portable freezer.
l TruCool® Hinged Cryoboxes, pack of 50, green color (BioCi-
sion®; BCS-207G), for storage of RNAlater tubes and cryovials
(referenced below, in Subheading 2.4).
2.4 Supragingival l Sterile wooden toothpicks, two per pack (DeRoyal; product
Biofilm Collection no. 30-413).
Supplies l RNAlater 1.5 mL TissueProtect Tubes (QIAGEN; product
no. 196594) for storage of biofilm samples to be carried forward
to metagenomics and metatranscriptomics analyses.
l TruCool™ cryogenic vials, sterile, 1 mL, self-standing, external
threads (BioCision®; BSC-2517) for downstream metabolomics
analyses.
2.5 Materials Used l Total RNA Purification Kit, 96-Well Format (Norgen Biotek,
for Nucleic Acid Corp.; P/N 24304, L/N 591338) which includes:
Purification – Robust lysis (RL) buffer (P/N 90055, L/N 589694).
– Wash solution A (P/N 90028, L/N 590111).
– 96-well filter plate (P/N 24304, L/N 591338).
– 96-well collection plate (P/N 24309).
– 96-well elution plate (P/N 24310).
l 96-well bead plate (Norgen Biotek, Corp.; P/N 65700, L/N
591926) kit, which includes:
– 96-well transfer plate (P/N 65702).
– 96-well bead plate (P/N 65703, L/N 591924).
l Beta-Mercaptoethanol (B-ME), CAS [60-24-2]
(MP Biomedicals, LLC; P/N 194834, L/N QR13543).
l Ethanol 200 Proof, anhydrous, CAS [64-17-5] (Decon Labs,
Inc.; P/N 2716, L/N DSP-MD.43).
l Molecular Biology Grade Water, sterile, CAS [7732-18-5]
(Corning; P/N 46-000-CM, L/N 14917003).
2.6 Software Used l Illumina bcl2fastq [18].

for Metagenomics and l FastQC [19].
Metatranscriptomics l bowtie2 [20].
Analyses
l vsearch [21].
l HUMAnN2 [22].
l Pathoscope2 [23].
l LEfSe [24].
l ALDEx2 R package [25].
l MetaPhlAn2 [26].
l KneadData (http://huttenhower.sph.harvard.edu/kneaddata).
l MaAsLin (https://huttenhower.sph.harvard.edu/maaslin).
l GraPhlAn (https://huttenhower.sph.harvard.edu/graphlan).
l ShotMAP [27].
Figure 1 illustrates basic small equipment (excluding CoolBoxes)
and supplies needed for supragingival biofilm collection. Figure 2
illustrates an assembled CoolBox unit with CFT 24 CoolRack and a
TruCool Cryogenic vial in place. Additional supplies and small
equipment assemblies are shown in the supplementary material
(Supplementary Material).
Fig. 1 Illustration of the supragingival plaque collection equipment and supplies:

Two packs of sterile toothpicks (pack of two toothpicks), presented from each
other side (DeRoyal. Box of 50. Product No. 30-413); one toothpick is used for
the collection of metagenomics/transcriptomics sample (collected from the
upper right quadrant), and the second toothpick for the metabolomics sample
(collected from the upper left quadrant). The toothpick with the metagenomics/
transcriptomics plaque sample is placed in a labeled RNAlater TissueProtect
Tube, containing 1.5 mL of RNAlater, and placed in a CoolRack XT M24 module
(left); the toothpicks with the metabolomics plaque sample is placed in a TruCool
cryogenic vial, labeled and immediately stored in a CoolRack CFT24 module
(right). Note: For additional information please see the Supplemental material
3 Methods
In the parent study, research examination teams use mobile dental

equipment, instruments, and supplies to set up clinical examination
stations where there is suitable space and a power outlet in commu-
nity locations (e.g., classrooms). Examiners are trained in the use of
basic child behavior guidance and management techniques rou-
tinely used in pediatric dentistry. The examination sequence includ-
ing biofilm collection takes place before or at least 30 (preferably
60) minutes after children have had breakfast or snack. The families
(and teachers, where applicable) are instructed not to brush the
children’s teeth the morning of the examination—the examiner
cleans the teeth with a toothbrush and dental floss after biospeci-
men collection. At the end of the day, the biofilm samples are stored
Fig. 2 An illustration of a sample stored in portable freezing storage container

(BioCision CoolBox XT CryoTube 24 Workstation, BioCision Catalogue No.
BSC-575). Note: For additional information please see the Supplemental
material
in a 20 C freezer until they are transferred to a core facility that is

equipped with 80 C freezers. The procedures are carried out as
follows:
3.1 Preclinical 1. Greet the child addressing them by their name and ideally
Procedures and kneeling down to their eye level.
Assessments 2. Explain the procedures that will follow in an age-appropriate
manner (e.g., “counting teeth”).
3.2 Supragingival 1. Collect supragingival biofilm (plaque) samples using the sterile
Biofilm Collection and toothpicks in a semi-reclined position, ideally a dental chair.
Storage Procedures Break off a small piece (~8–12 mm, or 1/300 –1/200 ) of each
toothpick and use to collect the biofilm.
2. One sample (biofilm sample A) from the upper right dentition
(facial surfaces of #A, #B, #C, #D, and #E according to the
universal nomenclature system or #55, #54, #53, #52, and #51
according to the FDI system) and another sample (biofilm
sample B) from the upper left dentition (facial surfaces of #F,
#G, #H, #I, and #J or #61, #61, #63, #64, and #61).
3. Place biofilm sample A (part of toothpick with metagenomics/
metatranscriptomics sample) in an RNAlater TissueProtect
1.5 mL tube and biofilm sample B (part of toothpick with

metabolomics sample) in a cryovial.
4. Add barcode or any other identifying labels on both vials.
5. Store in CoolRacks/CoolBoxes and close lid hermetically.
6. Add barcode or any other identifying label or notation on a
sample manifest that will accompany the stored vials.
3.3 Nucleic Acid Our group has optimized and tested both manual and high-
Purification Protocols throughput protocols for the processing of supragingival biofilm
among child research participants. We have used protocol Subhead-
ing 3.3.1 below to conduct metagenomics and transcriptomics
analyses among 118 study participants [12]—for that work, we
prioritized samples yielding 1 μg of total nucleic acid (NA) and
carried them forward to WGS and RNAseq analyses, obtaining on
average more than six million high-quality reads per sample and
informative results. The high-throughput protocol Subheading
3.3.2 is currently being optimized and tested in a similar fashion.
In general, we strongly recommend that nucleic acid protocols are
pilot-tested and validated, using clinical collection, biofilm speci-
men transport and storage, and analytical procedures identical to
the study conditions.
3.3.1 Sample 1. Inspect the sample vials and ensure integrity of the sample
Preparation for NA vessel. If you notice cracks on the sample vial, transfer the
Extraction sample to a DNase- and RNase-free screw cap vial that has
been properly labeled.
2. Thaw plaque samples at 37 C (~5 min in water bath, ~10 min
in heating block), and vortex for 5 min (use Mo Bio vortex
adapter or horizontal vortex mixer).
3. Spin tubes for 15 min at 14,000 g (13,200 g).
4. Using clean forceps, transfer the toothpick fragment into the
bead tube (pointed end first). Importantly, make note of
plaque visibility on toothpick fragment. Always clean forceps
between toothpick fragment transfers.
5. Remove most, if not all, of the RNAlater solution, without
disturbing the pellet. A small amount of liquid, ~50–100 μL,
can be left behind. Make notes as to the size of the pelleted
material.
6. Begin processing the pellet from the RNAlater TissueProtect
tube according to the manufacturer’s instructions as described
in Subheading 3.3.2.
We have determined that, using the protocols detailed in this
paper, the sequential processing of the nucleic acid sample is more
efficient in terms of yield than splitting the sample into DNA and
RNA using the AllPrep DNA/RNA Mini Kit (QIAGEN; catalogue
no. 80204).
3.3.2 Manual Total NA We carry out the manual purification protocol (Mo Bio Power-
Purification Biofilm RNA kit) according to the manufacturer’s protocol with
minor modifications, as follows:
(a) Thaw samples at 37 C for 10 min.
(b) Shake the samples at maximum speed for 5 min on a horizon-
tal vortex mixer.
(c) Centrifuge the plaque pellets and toothpick fragments (used
for collection) at 14,000 g for 15 min.
(d) Make note of plaque deposit on the toothpick fragment (not
visible, barely visible, visible, or conspicuous; see Supplemental
material for examples).
(e) Transfer the toothpick fragment into the bead tube.
(f) Follow the protocol according to the manufacturer’s instruc-
tions through step 15.
(g) Carry out steps involving disruption of the biofilm in the
presence of both the dissolved pellet from the RNAlater col-
lection tube and the toothpick fragment to optimize yield and
minimize material (biofilm) loss from what is remaining on the
toothpick.
(h) Omit all extraction steps dealing with DNA removal.
(i) Conclude elution of a total NA preparation with steps 21–26
of the manufacturer’s protocol.
3.3.3 High-Throughput In this protocol, we use the Total RNA Purification Kit, 96-Well
Protocol Format (Norgen Biotek, Corp.; catalogue no. 24304) and 96-Well
Bead Plate Kit (Norgen Biotek, Corp.; catalogue no. 65700). We
designed the protocol in a collaborative effort with the kit supplier.
The high-throughput total nucleic acid isolation method was opti-
mized and validated. The resulting procedure is as follows:
1. Thaw samples at 37 C for 10 min.
2. Shake the samples at maximum speed for 5 min on a horizontal
vortex mixer.
3. Centrifuge the plaque pellets and toothpick fragments (used
for collection) at 14,000 g for 15 min.
4. Make note of plaque deposit on the toothpick fragment (not
visible, barely visible, visible, or conspicuous; see Supplemental
material for examples).
5. Transfer the toothpick fragment into the bead tube.
6. Make note of plaque pellet in the sample plate (not visible,
barely visible, visible, or conspicuous).
7. Remove RNAlater solution.
(a) If pellet is not loose, decant the RNAlater solution.
(b) If the pellet is loose, remove most of the RNAlater solu-

tion by pipetting the liquid out of the tube, leaving up to
100 μL of solution.
8. Prepare robust lysis (RL) buffer with beta-mercaptoethanol
(B-ME), CAS [60-24-2] (MP Biomedicals, LLC; catalogue
no. 194834) (RL/BME) at a ratio of 10 μL of BME per
1 mL of RL buffer (1:100 ratio).
(a) Prepare enough RL/BME solution to treat all samples
been processed and an additional sample (n + 1). For
example, to process eight samples, prepare 8 + 1 ¼ 9, as
follows:
Prepare RL/BME j9 400 μL ¼ 3600 μL; 36 μL
β-ME + 3.6 mL RL Buffer
(b) Use 400 μL of RL/BME solution per sample.
(c) If processing a small number of samples, n < 5, prepare
enough solution for n + 0.5 reactions.
9. Resuspend each plaque pellet in 400 μL RL/ME solution.
10. Transfer the sample in RL/ME solution to a well of a bead
plate and secure the silicone mat. Make note of the wells used
and sample order.
11. Vortex briefly to mix.
12. Incubate the sample at 65 C for 5 min, inverting every 2 min
to mix. Secure the silicone mat with a piece of reinforcement
foam, an assembled rounded-bottom 96-well plate and lid, and
clamps on all four corners.
(a) Clamps and reinforcement material are not required if a
vertical plate homogenizer is used.
13. Vortex the sample at maximum speed for 5 min on a horizontal
vortex mixer (Fisher Scientific multi-tube vortexer). Disassem-
ble the reinforcement foam, 96-well plate and lid, and clamps.
14. Centrifuge the bead plate at 3000 g for 3 min at room
temperature.
15. Dispense 200 μL of absolute (100%) ethanol into each well of a
clean ribonuclease-free collection microplate (96-Well Bead
Plate Kit).
16. Collect the supernatants and transfer into the wells of the
ethanol-containing collection microplate. Make note of the
wells used and sample order.
17. Mix each lysate and ethanol by gently pipetting up and down
3–5 times.
18. Bind the nucleic acids to the filter plate as follows:
(a) Place the 96-well filter plate on top of a provided 96-well
collection plate.
(b) Apply up to 500 μL of the lysate with the ethanol into

each well of the 96-well filter plate.
(c) Centrifuge the assembly at maximum speed or 3000 g
(~3000 rpm) for 2 min.
(d) Discard the flow-through. Reassemble the 96-well filter
plate and the bottom plate. Ensure that the lysate from
each well has passed through into the bottom plate. If the
entire lysate volume has not passed, centrifuge for an
additional 2 min.
(e) Run all the samples through the spin column.
19. Wash the nucleic acids thrice (3) as follows:
(a) Apply 400 μL of Wash Buffer A.
(b) Centrifuge the assembly at maximum speed or 3000 g
(c) Discard the flow-through.
20. Pat the bottom of the 96-well filter plate dry with a clean
lab wipe.
21. Reassemble the 96-well filter plate and the bottom plate.
22. Centrifuge the assembly at maximum speed or 3000 g
(~3000 rpm) for 5 min to completely dry the plate.
23. Elute the nucleic acids into a new nuclease-free tube by adding
60 μL of pre-warmed nuclease-free water (65 C) directly onto
the center of the spin column filter to ensure complete elution
of nucleic acids.
24. Incubate at room temperature for 1 min.
25. Centrifuge the assembly at maximum speed or 3000 g
26. Record the volume of the elute (usually ~50 μL).
27. Transfer sample into a labeled nuclease-free screw cap tube
with cap.
28. Proceed to nucleic acid quantitation and QA.
3.4 Whole Genome To optimize results, each DNA sample should generally meet the
Shotgun (WGS) following standard requirements:
Sequencing: DNA (a) Has not undergone multiple freeze-thaw cycles as they can
Requirements, Library lead to DNA damage.
Preparation, and
(b) Has not been exposed to high temperatures (e.g., >65 C for
Sequencing
1 h can cause a detectable decrease in sequence quality) or pH
3.4.1 Several Factors extremes (<6 or >9).
Can Influence DNA Quality (c) Has an OD260/OD280 ratio of 1.8–2.0.
and, Thus, Read Length
and Quality of Sequencing
(d) Does not contain RNA contamination.
(e) Does not contain denaturants (e.g., guanidinium salts or phe-

nol) or detergents (e.g., SDS or Triton X100).
(f) Does not contain carryover contamination.
Dilute DNA in DNase-free water or low-salt, weakly buffered
solutions containing little or no metal ion chelating agents such as
EDTA (e.g., 10 mM Tris). The minimum amount of DNA per
sample required for WGS library preparation as described in Sub-
heading 3.4.1 is 1 ng.
3.4.2 Process 1 ng of 1. Quantify the sample DNA using PicoGreen reagent and opti-
Intact Genomic DNA Using mize concentration to 0.2 ng/μL.
the Nextera XT DNA 2. Using Nextera XT transposome, simultaneously fragment the
Sample Preparation Kit input DNA and add platform-specific adapter sequences:
(Illumina, San Diego, CA)
(a) Label a new 96-well plate.
(b) Add 10 μL of Tagment DNA Buffer to each well to be
used in this assay. Change tips between samples.
(c) Add 5 μL of input DNA at 0.2 ng/μL (1 ng total) to each
sample well of the plate.
(d) Add 5 μL of Amplicon Tagment Mix to the wells contain-
ing input DNA and buffer. Change tips between samples.
(e) Using a multichannel pipette, gently pipette up and down
five times to mix. Change tips between samples.
(f) Cover the NTA plate with Microseal.
(g) Centrifuge at 280 g at 20 C for 1 min.
(h) Place the NTA plate in a thermocycler and run the follow-
ing program:
l 55 C for 5 min.
l Hold at 10 C.
3. Once the sample reaches 10 C, immediately add 5 μL of
neutralization buffer to each well of the plate. Change tips
between samples.
4. Amplify tagmented DNA via a limited-cycle PCR program
adding index 1(i7) and index 2(i5) (Illumina) in unique com-
bination for each sample, as well as primer sequences required
for cluster formation:
(a) Add 15 μL of Nextera PCR Master Mix to each well of the
plate containing index primers. Change tips between
samples.
(b) Using a multichannel pipette, add 5 μL of index 2 primers
(white caps) to each column of the plate. Changing tips
between columns is required to avoid cross
contamination.
(c) Using a multichannel pipette, add 5 μL of index 1 primers

(orange caps) to each row of the plate. Tips must be
changed after each row to avoid index cross
contamination.
(d) Using a multichannel pipette, gently pipette up and down
3–5 times to mix.
(e) Cover the plate with Microseal and seal with a rubber
roller.
(f) Centrifuge at 280 g at 20 C for 1 min.
(g) Perform PCR using the following program on a thermal
cycler:
l 72 C for 3 min.
l 95 C for 30 s.
l 12 cycles of:
– 95 C for 10 s.
– 55 C for 30 s.
– 72 C for 30 s.
l 72 C for 5 min.
l Hold at 10 C.
3.4.3 Purify Each Library This step allows to purify the library DNA and provides a size
Using Agencourt® selection step that removes very short library fragments from the
AMPure® XP Reagent population:
(Beckman Coulter,
1. Centrifuge the plate at 280 g for 1 min (20 C) to collect
Brea, CA)
condensation.
2. Vortex the AMPure XP beads for 30 s to ensure that the beads
are evenly dispersed. Add an appropriate volume of beads to a
trough.
3. Using a multichannel pipette, add 30 μL of AMPure XP beads
to each well of the plate and pipette mix up and down 10 times.
4. Incubate at room temperature without shaking for 5 min.
5. Place the plate on a magnetic stand for 2 min or until the
supernatant has cleared and remove and discard the
supernatant.
6. With the plate on the magnetic stand, wash the beads twice
with freshly prepared 80% ethanol and allow the beads to
air-dry for 15 min.
7. Remove the plate from the magnetic stand. Using a multichan-
nel pipette, add 25 μL of resuspension buffer to each well.
8. Gently pipette mix up and down 10 times, changing tips after
each column and incubate at room temperature for 2 min.
9. Place the plate on the magnetic stand for 2 min or until the
supernatant has cleared.
10. Using a multichannel pipette, carefully transfer 23 μL of the
supernatant to a new plate. Change tips between samples to
avoid cross contamination.
3.4.4 Quantify Each Reagent (Molecular Probes, Thermo Fisher Scientific division,
Library Using Quant-iT™ Waltham, MA) according to manufacturer instructions.
PicoGreen® dsDNA
3.4.5 Calculate and Pool

All Libraries in Equimolar
Amounts to One Tube
3.4.6 Follow the HiSeq For preparation of the library to 10 pM for loading:
2500 System Guide
1. Heat denature the resulting pool before loading on the HiSeq
Protocol (Illumina)
reagent cartridge and on the HiSeq 2500 instrument (Illu-
mina). Denatured PhiX library can be used as an internal
control or to balance low-diversity libraries. For libraries with
low complexity, combine 30% PhiX with your diluted sample.
2. Enter run parameters according to consumable ID, indexing
option, and flow cell used.
If edited properly, for dual-indexed PE250 runs, the instru-
ment will indicate that the run is 250 8 8 250. The run
takes approximately 1 week to complete.
3.5 RNA Sequencing (a) Suspend RNA samples in RNase-free water or 1 TE buffer
(RNAseq): RNA prepared with RNase-free water.
Preparation, (b) Assess RNA integrity using the Agilent Bioanalyzer (or any
Requirements, and similar system), with optimal RNA quality number (RQN) or
Library Preparation RNA integrity number (RIN) being 8.
3.5.1 RNA Preparation (c) We recommend a DNase treatment step in the RNA isolation
and Requirements protocol.
(d) We recommend using fluorometric methods such as Quant-
iT™ RiboGreen® for RNA quantification.
500 ng of total RNA is required for RNA library preparation
according to the procedures described in Subheading 3.5.2.
3.5.2 Library Preparation 1. From a sample of total RNA, remove noncoding rRNA using
Illumina Ribo-Zero Epidemiology Kit (San Diego, CA).
(a) Wash the rRNA-specific magnetic beads off the storage
buffer.
(b) Mix with 500 ng of total sample RNA with rRNA removal
solution and incubate for 10 min at 65 C.
(c) Add pre-washed rRNA-specific magnetic beads from step

(a) and incubate for 5 min at 50 C.
(d) Place samples on magnetic stand for 15 min in room
temperature.
(e) Aspirate the supernatant containing coding RNA and
proceed to library preparation.
2. mRNA library preparation using TruSeq® Stranded mRNA
Sample Preparation Kit (San Diego, CA).
(a) Fragment and prime mRNA using divalent cation
priming solution at 94 C for 8 min.
(b) Equilibrate first strand synthesis Actinomycin D mix
(FSS) to room temperature.
(c) Add FSS to mRNA samples.
(d) Place samples on the thermal cycler and use the following
thermal conditions:
l 25 C for 10 min.
l 42 C for 15 min.
l 70 C for 15 min.
l Hold at 4 C.
(e) Remove samples from thermal cycler.
(f) Add second strand marking master mix (SMM) to each
sample.
(g) Incubate at 16 C for 1 h.
(h) Purify double-stranded cDNA using Agencourt®
AMPure® XP Reagent (Beckman Coulter, Brea, CA).
(i) Adenylate 30 ends of cDNA by adding A-tailing mix to
each sample and incubating in the following thermal
conditions:
l 37 C for 30 min.
l 70 C for 5 min.
l Hold at 4 C.
(j) Ligate adapters by adding ligation mix and adapters, each
with a unique barcode.
(k) Incubate samples at 30 C for 10 min.
(l) Add stop solution to quench ligation reaction.
(m) Purify cDNA libraries using Agencourt® AMPure® XP
Reagent.
(n) Enrich DNA fragments by adding PCR primer cocktail
and master mix to each sample and incubating in the
following thermal conditions for 15 cycles:
l 98 C for 10 s
l 60 C for 30 s
l 72 C for 30 s
(o) Carry out a final purification of cDNA libraries using
Agencourt® AMPure® XP Reagent.
3. Validate quality of cDNA libraries on Agilent TapeStation.
4. Assess library concentration using Quant-iT PicoGreen
dsDNA Reagent from Thermo Fisher Scientific (Eugene, OR).
5. Pool all libraries in equimolar amounts.
3.6 Analysis 1. De-multiplex and convert to FASTQ format paired-end

Methods for sequencing run results using bcl2fastq [12].
Metagenomics and 2. Assess quality of sequencing results with FastQC [13].
Metatranscriptomics 3. Align paired-end reads against the human Hg19 reference
(Bacterial using bowtie2 [14]. Retain reads that do not align.
Transcriptomics)
4. Join paired-end reads that do not align to reference into single-
end reads with vsearch [15] when possible. Retain both reads
that join as well as reads that do not.
5. Combine all reads, joined and unjoined, into a single single-
end data set. Align against the human hg19 reference using
bowtie2 [14]. Retain all reads that do not align.
6. Perform microbial community profiling using
MetaPhlAn2.2 [20].
7. Classify all retained reads by taxonomy, gene family, path cov-
erage, and path abundance using HUMAnN2 [16].
8. Classify all retained reads by taxonomy using
Pathoscope2 [17].
9. Profile strain-level variation in microbial communities using
StrainPhlAn, as referenced in Subheading 2.6.
10. Identify differentially abundant taxa with LEfSe [18], ALDEx2
[19], or MaAsLin, as referenced in Subheading 2.6.
11. Produce taxonomic or phylogenetic representations (i.e., cir-
cular heat maps and bar plots) of analytical results using GraPh-
lAn, as referenced in Subheading 2.6.
3.7 Outline of The choice of statistical methods used for association analyses
Statistical Methods between bacterial taxa or transcripts and health/disease statuses is
Used to Identify driven by the distributional assumptions and characteristics of these
Differentially data. These analyses are, by nature, subject to multiplicity (i.e.,
Abundant Taxa or multiple testing) issues.
Differentially
Expressed Bacterial
Genes
3.7.1 Modeling Taxa or gene abundance or expression levels are represented as

Assumptions counts, ratios, or proportions: normal, binomial, and beta are the
most widely used distributions. In general, it is expected that
models that can handle zero inflation will yield higher power com-
pared to their non-zero-inflated counterparts, due to the high
prevalence of excess zeros in microbiome data. In the context of
moderately small numbers of zeros, models that do not explicitly
address zero inflation can also be used. The following models and
procedures can be considered for testing such associations with
disease status.
i. MAST [28] Zero-inflated log-normal distribution

ii. Two-part logistic beta model Bernoulli and beta distribution
[29, 30]
iii. Two-part Wilcoxon rank sum Bernoulli and normal (rank)
test [31]
iv. Zero-inflated negative binomial Zero-inflated negative binomial
model [32, 33]
v. LEfSe [24] Normal
vi. Kruskal-Wallis test [34] Normal (rank)
Models i, ii, iv, and v can incorporate continuous and/or

discrete covariates in the model, while models iii and vi can be
modified to handle discrete covariates (e.g., in the context of
batch effects). When the disease status of interest is expressed as a
multi-category variable, models i, ii, iv, and v are applicable,
whereas models iii and vi are not. Of note, models i, ii, iii, and iv
are zero-inflated models.
3.7.2 Multiple Testing Association analyses between bacterial taxa (or transcripts) and
and Control of Type-I Error health/disease statuses are usually performed at the individual
taxon level. Due to the high number of taxa to be tested, a typical
multiple comparisons problem arises. To control type-I error, false
discovery rate (FDR) controlling procedures and familywise error
rate (FWER) controlling procedures can be used to correct/adjust
p-values. FWER procedures include, among others, the Bonferroni
and Holm corrections; these provide relatively stringent control of
type-I error and are less powered than FDR. Typical FDR proce-
dures include the Benjamini-Hochberg (BH) and Benjamini-Yeku-
tieli (BY) corrections. Both FWER and FDR can be implemented
using the R function p.adjust (https://stat.ethz.ch/R-manual/R-
devel/library/stats/html/p.adjust.html).
Although the Bonferroni correction and BH procedure are
popular, they both assume that the individual tests are independent.
In the context of phylogenetic structure, which is present de facto
in microbiome analyses, the hierarchical Benjamini-Hochberg
procedure (HBH) [35] can be used to construct a two-stage frame-

work. At the first stage, screening tests are performed on higher-
rank taxonomic groups (e.g., families) and at the second stage,
lower-rank taxonomic groups (e.g., species) are tested with BH
procedures [36].
3.8 Sample Metabolon has designed and implemented a global metabolomics

Processing and platform aimed to overcome known challenges associated with
Metabolomics broad-range metabolite profiling. These advanced analytical meth-
Analyses of the odologies allow for the detection of metabolites in all major metab-
Supragingival Biofilm olite classes, an analytical feat due to the vast differences in
Using the Metabolon molecular size, physical and chemical properties, and physiological
DiscoveryHD4™ concentrations across classes. The procedures described can effi-
Platform ciently detect molecules ranging from small highly polar com-
pounds, like TCA cycle intermediates, to large hydrophobic
complex lipids with a single sample aliquot. The Metabolon pipe-
line currently used is the proprietary DiscoveryHD4™ platform
[13–17], including multiple mass spectrometry methods, a large
reference library of authenticated metabolite standards and a suite
of patented informatics and quality control software. We present
laboratory steps to process the supragingival biofilm samples for
metabolomics analyses in Subheading 3.8.1.
3.8.1 Sample Keep samples stored at 80 C until needed and then thaw them
Preparation on ice just prior to extraction. Supragingival biofilm samples col-
lected and stored with a toothpick are extracted using an intact
biopsy sample extraction method as described below:
(a) Prepare the extraction solvent by adding 40 mL of HPLC
grade methanol (CAS 67-56-1) and 10 mL of ultrapure
water to a 50 mL conical tube. Invert several times to mix.
[Note: The methanol contains four recovery standards, DL-2-
fluorophenylglycine, tridecanoic acid, d6-cholesterol, and
4-chlorophenylalanine, to allow confirmation of extraction
efficiency.]
(b) Add 1 mL of extraction solvent to each pre-labeled sample vial
for your sample set.
(c) Deposit each biofilm sample on a toothpick into the
corresponding pre-labeled sample vial containing room tem-
perature (RT) extraction solvent (one biopsy per vial). The
sample will come off of the toothpick with gentle swirling in
the MeOH.
(d) Cap vials and incubate samples in extraction solvent at room
temperature, on the benchtop, for a minimum of 4 h and up to
48 h. No agitation is necessary. [Note: While samples may be
incubated in extraction solvent for 4 h to 48 h, it is critical that
all samples in a study are incubated for the same amount of

time.]
(e) Following incubation, remove the toothpick from the extrac-
tion solvent using forceps. It is crucial to rinse forceps in a
rinse solution (80% MeOH, 20% water) between each tooth-
pick retrieval.
(f) Tighten the lids on the vials and the MeOH extracted meta-
bolites are ready to be analyzed by UPLC.
(g) Take four aliquots of each sample from the methanol extract
and dry them.
(h) Reconstitute two aliquots of each sample in 50 μL of 6.5 mM
ammonium bicarbonate in water (pH 8) for the negative ion
analysis.
(i) Reconstitute two more aliquots of each sample using 50 μL of
0.1% formic acid in water (pH ~3.5) for the positive ion
analysis. [Note: Reconstitution solvents contain instrument
internal standards to assess instrument performance and to
serve as retention index markers for chromatographic
alignment.]
3.8.2 Ultrahigh- (a) Conduct UPLC separation using a Waters Acquity UPLC
Performance Liquid (Waters, Milford, MA).
Chromatography-Tandem (b) Use the mobile phase consisting of 0.1% formic acid in water
Mass Spectroscopy (UPLC- (A) and 0.1% formic acid in methanol (B) for reverse-phase
MS/MS) (RP) positive ion analysis.
(c) Use the mobile phase consisting of 6.5 mM ammonium bicar-
bonate in water, pH 8 (A), and 6.5 mM ammonium bicarbon-
ate in 95% methanol/ 5% water for the reverse-phase negative
ion analysis. Use a sample injection volume of 5 μL and a
2 needle loop overfill. Use a separate acid and base-dedicated
2.1 mm 100 mm Waters BEH C18 1.7 μm columns held at
40 C for separations.
(d) Use the mobile phase consisting of 10 mM ammonium for-
mate in 15% water, 5% methanol, 80% acetonitrile (effective
pH 10.16 with NH4OH) (A) and 10 mM ammonium formate
in 50% water, 50% acetonitrile (effective pH 10.60 with
NH4OH) (B) for hydrophilic interaction liquid chromatogra-
phy (HILIC). Use the same sample injection volume as in the
RP method. The stationary phase consists of a
2.1 mm 150 mm Waters BEH Amide 1.7 μm column held
at 40 C.
(e) Conduct MS analysis alternating between MS and data-
dependent MS scans using dynamic exclusion. The scan
range varies slightly between methods but covers
70–1000 m/z.
(f) Archive the raw data files before carrying forward to analysis.
An informatics pipeline is described in Subheading 3.8.3.
3.8.3 The Metabolon® The informatics system comprises a Laboratory Information Man-
Informatics Pipeline for agement System (LIMS), data extraction and peak-identification
Data Extraction, Compound software, data processing tools for QC and compound identifica-
Identification, Quality tion, and libraries for interpretation and visualization tools. Execute
Control (QC), the informatics pipeline as follows:
Normalization, and
(a) Extract, peak-identify, and QC process the raw data using the
Interpretation
Metabolon® software pipeline on a research computing server.
(b) Identify compounds by comparison to library entries of pur-
ified standards or recurrent unknown entities, based on
authenticated standards of retention time/index (RI), mass
to charge ratio (m/z), and chromatographic data (including
MS/MS spectral data) on all molecules present in the library.
(c) Identify biochemical compounds based on three criteria:
(1) retention index within a narrow RI window of the pro-
posed identification, (2) accurate mass match to the library
10 ppm, and (3) the MS/MS forward and reverse scores
between the experimental data and authenticated standards.
The MS/MS scores are derived based on a comparison of the
ions present in the experimental spectrum to the ions present
in the library spectrum.
(d) Perform QC procedures (e.g., checks of consistency of peak
identification between samples and library matches for each
compound) as a means of improving identification of true
chemical entities and removing system artifacts, misassign-
ments, and background noise.
(e) Quantify peaks for each metabolite using area under the curve.
(f) Include a data normalization step to correct variation resulting
from instrument inter-day tuning differences for studies span-
ning multiple days. Essentially, correct each compound in
run-day blocks by registering the medians to equal one
(1.00) and normalizing each data point proportionately.
4 Notes
1. When feasible, obtaining tooth surface-specific biofilm samples

is in principle preferable than collecting pooled samples—tooth
surface-specific samples can be more informative as they can be
linked to localized disease (e.g., a specific caries lesion) [5] or
other anatomical or ecological features. Of note, the protocol
does not describe biofilm collection from occlusal tooth sur-
faces, which are systematically different than facial/buccal
surfaces of the maxillary dentition and perhaps most informa-

tive in some studies or for some analyses.
2. Adherence to instructions (e.g., not brushing prior to biofilm
collection) is an important factor affecting the amount (and
potentially the quality) of biofilm collected for analysis.
3. Efforts to collect a “full thickness” biofilm sample are well-
invested—arguably, the biofilm closest to the tooth surface is at
least equally informative for ECC than the outmost layer. We
recommend a sweeping movement with the toothpick from the
distal to the mesial of each facial tooth surface, while staying
supragingivally at all times.
4. If a sterile curette can be used instead of toothpicks for supra-
gingival biofilm collection, it may be preferable, as it would
eliminate the need to separate biofilm material from the tooth-
pick in later steps—in our case we chose sterile toothpicks for
practical reasons, due to the conduct of the study in the field.
5. Quality assessment and validation studies are highly recom-
mended to ensure that collection methods, cold chain, nucleic
acid purification methods, and all other processes are working
as expected to produce the desired downstream ‘omics data.
6. During the manual nucleic acid purification process using the
Mo Bio kit, we have used the Mo Bio vortex adapter and a
Fisher horizontal vortex mixer (CAT #02215450) inter-
changeably, with no differences in NA yield or quality.
7. WGS is a versatile method that allows taxonomic profiling of
microbial communities to species or, when utilizing additional
tools, down to the strain level—it also enables functional
profiling of metagenomic and metatranscriptomic sequence
data, including aggregated whole-community level pathway
reconstruction.
8. Metatranscriptomic sequence data are a logical and informative
complement to metagenomics because they can provide infor-
mation regarding the functional activity of the identified micro-
bial communities at different taxonomic levels. Importantly,
they may help illuminate ecological units with high transcrip-
tional activity that might otherwise not be identified or appre-
ciated from a taxonomic standpoint, due to rareness.
9. Segata and colleagues [37] offer an excellent review of techno-
logical and computational approaches available for microbiome
research. The development of analytical tools for metage-
nomics and metatranscriptomics analyses is rapidly evolving—
association analyses of microbial communities at different levels
and respective functional profiles can be performed by stand-
alone software specifically developed for this purpose (e.g.,
MaAsLin, LEfSe, ALDEx2) or using individualized approaches
in standard statistical environments (e.g., R) and universal

packages (e.g., lm, glm, glmm, etc.).
10. Handling and processing of WGS data is time and computa-
tionally intensive, involving numerous steps. We recommend
that WGS analysis pipelines are well-documented and fully
automatized or divided into broad processes (e.g., KneadData
tool for quality control) to promote efficiency and reproduc-
ibility of multistage cascades of bioinformatics tools.
11. We recommend the consideration of cloud-based computing
solutions as a means to accommodate the ever-increasing
computational demands of WGS analyses in large-scale studies
or in settings where advanced, resource-intensive statistical
methods are used, while balancing cost.
12. In metabolomics analyses, there can be similarities between
biochemical compounds based on RI index, mass match, and
MS/MS scores—the simultaneous use of all three estimates
helps distinguish and differentiate these molecules. Currently,
there are over 3300 commercially purified standard com-
pounds available into Metabolon’s LIMS. Additional mass
spectral entries have been created for structurally unnamed
biochemicals, which have been identified by virtue of their
recurrent nature, both chromatographic and mass spectral.
These compounds have the potential to be identified in the
future.
13. We recommend including several blanks (e.g., 10 toothpick
fragments without samples, handled according to study condi-
tions) in metabolomics analyses, to detect and account for any
resulting background noise due to this collection medium.
14. For metabolomics analyses that do not span more than 1 day,
no normalization is necessary, other than for purposes of data
visualization. Biochemical data can be normalized to an addi-
tional factor (e.g., cell counts, total protein as determined by
Bradford assay, osmolality, etc.) to account for differences in
metabolite levels due to differences in the amount of material
present in each sample.
Acknowledgments
This work was supported by a grant from the National Institutes of

U01-DE025046. DS is supported by the Swedish Research Coun-
cil (4.1-2016-00416). The Microbiome Core is supported in part
by the NIH/National Institute of Diabetes and Digestive and
Kidney Diseases grant P30 DK34987.
References
1. Nascimento MM, Zaura E, Mira A, (2018) Metagenomics of early childhood oral
Takahashi N, Ten Cate JM (2017) Second era health and early childhood caries. J Dent Res
of OMICS in caries research: moving past the 97 (Spec Iss A):2412231 (CADR/AADR)
phase of disillusionment. J Dent Res 96 13. Evans AM, Bridgewater BR, Liu Q, Mitchell
(7):733–740 MW, Robinson RJ, Dai H, Stewart SJ, DeHa-
2. Divaris K (2016) Predicting dental caries out- ven CD, Miller LA (2014) High resolution
comes in children: a “risky” concept. J Dent mass spectrometry improves data quantity and
Res 95(3):248–254 quality as compared to unit mass resolution
3. Ballantine J, Carlson JC, Zandona A, Agler CS, mass spectrometry in high-throughput
Zeldin LP, Rozier G, Roberts MW, Basta PV, profiling metabolomics. Metabolomics 4(2):1
Luo J, Antonio-Obese ME, McNeil D, 14. Evans AM, Mitchell MW, Dai H, DeHaven CD
Weyant R, Crout RJ, Slayton R, Levy S, Shaffer (2012) Categorizing ion-features in liquid
JR, Marazita ML, North KE, Divaris K (2018) chromatography/mass spectrometry metabo-
Exploring the genomic basis of early childhood lomics data. Metabolomics 2(110):2153–0769
caries: a pilot study. Int J Paediatr Dent 28 15. DeHaven CD, Evans AM, Dai H, Lawton KA
(2):217–225 (2012) Software techniques for enabling high-
4. Kilian M, Chapple IL, Hannig M, Marsh PD, throughput analysis of metabolomic datasets.
Meuric V, Pedersen AM, Tonetti MS, Wade In: Metabolomics 2012. Intech, Vienna
WG, Zaura E (2016) The oral microbiome— 16. DeHaven CD, Evans AM, Dai H, Lawton KA
an update for oral healthcare professionals. Br (2010) Organization of GC/MS and LC/MS
Dent J 221(10):657–666 metabolomics data into chemical libraries. J
5. Nyvad B, Crielaard W, Mira A, Takahashi N, Chem 2(1):9
Beighton D (2013) Dental caries from a 17. Evans AM, DeHaven CD, Barrett T,
molecular microbiological perspective. Caries Mitchell M, Milgram E (2009) Integrated,
Res 47(2):89–102 nontargeted ultrahigh performance liquid
6. Featherstone JD (2004) The continuum of chromatography/electrospray ionization tan-
dental caries—evidence for a dynamic disease dem mass spectrometry platform for the iden-
process. J Dent Res 83 Spec No C:C39–C42 tification and relative quantification of the
7. Drury TF, Horowitz AM, Ismail AI, Maertens small-molecule complement of biological sys-
MP, Rozier RG, Selwitz RH (1999) Diagnos- tems. Anal Chem 81(16):6656–6667
ing and reporting early childhood caries for 18. Bcl2Fastq 2.20 (2017) Illumina, Inc. San
research purposes. A report of a workshop Diego, CA
sponsored by the National Institute of Dental 19. FastQC 0.11.5 (2016) Babraham Institute,
and Craniofacial Research, the Health Cambridge, UK
Resources and Services Administration, and 20. Langmead B, Salzberg SL (2012) Fast gapped-
the Health Care Financing Administration. J read alignment with Bowtie 2. Nat Methods 9
Public Health Dent 59(3):192–197 (4):357–359
8. Divaris K (2017) Precision dentistry in early 21. Rognes T, Flouri T, Nichols B, Quince C,
childhood: the central role of genomics. Dent Mahé F (2016) VSEARCH: a versatile open
Clin North Am 61(3):619–625 source tool for metagenomics. PeerJ 4:e2584
9. Tringe SG, von Mering C, Kobayashi A, Sala- 22. Abubucker S, Segata N, Goll J, Schubert AM,
mov AA, Chen K, Chang HW, Podar M, Short Izard J, Cantarel BL, Rodriguez-Mueller B,
JM, Mathur EJ, Detter JC, Bork P, Zucker J, Thiagarajan M, Henrissat B,
Hugenholtz P, Rubin EM (2005) Comparative White O, Kelley ST, Methé B, Schloss PD,
metagenomics of microbial communities. Sci- Gevers D, Mitreva M, Huttenhower C (2012)
ence 308(5721):554–557 Metabolic reconstruction for metagenomic
10. Gilbert JA, Hughes M (2011) Gene expression data and its application to the human micro-
profiling: metatranscriptomics. Methods Mol biome. PLoS Comput Biol 8(6):e1002358
Biol 733:195–205 23. Hong C, Manimaran S, Shen Y, Perez-Rogers
11. Holmes E, Wilson ID, Nicholson JK (2008) JF, Byrd AL, Castro-Nallar E, Crandall KA,
Metabolic phenotyping in health and disease. Johnson WE (2014) PathoScope 2.0: a com-
Cell 134(5):714–717 plete computational framework for strain iden-
12. Divaris K, Roach J, Basta PV, Ferreira Zandona tification in environmental or clinical
AG, Ginnis J, Meyer BD, Hu S, Simancas- sequencing samples. Microbiome 2:33
Pallares MA, Butz N, Azcarate-Peril MA
24. Segata N, Izard J, Waldron L, Gevers D, microbiome compositional data. PLoS Com-
Miropolsky L, Garrett WS, Huttenhower C put Biol 14(7):e1006329
(2011) Metagenomic biomarker discovery 31. Wagner BD, Robertson CE, Harris JK (2011)
and explanation. Genome Biol 12(6):R60 Application of two-part statistics for compari-
25. Fernandes AD, Macklaim JM, Linn TG, son of sequence variant counts. PLoS One 6
Reid G, Gloor GB (2013) ANOVA-like differ- (5):e20296
ential expression (ALDEx) analysis for mixed 32. Hu MC, Pavlicova M, Nunes EV (2011) Zero-
population RNA-Seq. PLoS One 8(7):e67019 inflated and hurdle models of count data with
26. Segata N, Waldron L, Ballarini A, extra zeros: examples from an HIV-risk reduc-
Narasimhan V, Jousson O, Huttenhower C tion intervention trial. Am J Drug Alcohol
(2012) Metagenomic microbial community Abuse 37(5):367–375
profiling using unique clade-specific marker 33. Preisser JS, Das K, Long DL, Divaris K (2016)
genes. Nat Methods 9(8):811–814 Marginalized zero-inflated negative binomial
27. Ghosh S, Chan CK (2016) Analysis of regression with application to dental caries.
RNA-Seq data using TopHat and cufflinks. Stat Med 35(10):1722–1735
Methods Mol Biol 1374:339–361 34. Kruskal WH, Wallis WA (1952) Use of ranks in
28. Finak G, McDavid A, Yajima M, Deng J, Ger- one-criterion variance analysis. J Am Stat Assoc
suk V, Shalek AK, Slichter CK, Miller HW, 47(260):583–621
McElrath MJ, Prlic M, Linsley PS (2015) 35. Yekutieli D (2008) Hierarchical false discovery
MAST: a flexible statistical framework for asses- rate–controlling methodology. J Am Stat Assoc
sing transcriptional changes and characterizing 103(481):309–316
heterogeneity in single-cell RNA sequencing 36. Hu J, Koh H, He L, Liu M, Blaser MJ, Li H
data. Genome Biol 16(1):278 (2018) A two-stage microbial association
29. Peng X, Li G, Liu Z (2016) Zero-inflated beta mapping framework with advanced FDR con-
regression for differential abundance analysis trol. Microbiome 6(1):131
with metagenomics data. J Comput Biol 23 37. Segata N, Boernigen D, Tickle TL, Morgan
(2):102–110 XC, Garrett WS, Huttenhower C (2013)
30. Chai H, Jiang H, Lin L, Liu L (2018) A mar- Computational meta’omics for microbial com-
ginalized two-part Beta regression model for munity studies. Mol Syst Biol 9:666
Chapter 41
Saliva and Gingival Crevicular Fluid (GCF) Collection

for Biomarker Screening
Petros Papagerakis, Li Zheng, Doohak Kim, Raed Said, Amber A. Ehlert,
Kevin K. M. Chung, and Silvana Papagerakis
Abstract
Assaying different biological markers (biomarkers) is commonly used to monitor health status and aid in the
diagnosis of diseases. With the recent advances in highly sensitive protein assays, whole saliva (WS) and
gingival crevicular fluid (GCF) appear to be fluids that may contain important biomarkers with various
applications in dentistry and medicine. Herein, we describe the process of GCF and WS sample collection
and preparation for assaying clinically relevant biomarkers in clinical screening trials. Analysis of biomarkers
in WS and GCF represents an easy and practical approach for the diagnosis and screening of different
pathological conditions particularly in epidemiological surveys.
Key words Biomarkers, Saliva, Gingival crevicular fluid, Screening, Diagnostics
1 Introduction
Biological markers (biomarkers) are used to objectively measure

normal biological processes, pathogenic processes, or responses to
a therapeutic intervention. In addition, biomarker screening can be
used to detect a person’s susceptibility to a specific disease or
disease process. Biomarkers have been traditionally analyzed in
body fluids such as blood, urine, and cerebrospinal fluid. With the
recent advances in highly sensitive protein assays, whole saliva
(WS) and gingival crevicular fluid (GCF) appear to be potential
fluids that may contain important biomarkers with various applica-
tions in dentistry and medicine [1–4]. One main advantage of
biomarker assays is their ability to detect changes before clinical
signs or symptoms become obvious. For example, analysis of
dentin- and cementum-specific biomarkers can be utilized to
detect orthodontic root resorption before radiographic signs are
noticed [5, 6].
549
550 Petros Papagerakis et al.
Whole saliva is the fluid that is found in the oral cavity. It

contains secretions from the major and minor salivary glands as
well as non-salivary components. These non-salivary components
include gingival crevicular fluid, nasal and bronchial secretions,
serum and blood derivatives from wounds, desquamated epithelial
linings, food components, and microorganisms that reside in the
oral cavity [7]. Gland-specific saliva can be collected to diagnose
pathology specific to one of the major salivary glands, whereas WS
is collected to reflect systemic conditions. Indeed, several types of
inflammatory biomarkers associated with both oral diseases and
systemic diseases have been detected in WS, such as interleukins
1β, 6, and 8 (IL-1β, IL-6, and IL-8), tumor necrosis factor-α
(TNF-α), and matrix metalloproteinases 8 and 9 (MMP-8
and MMP-9) [8–12]. Moreover, an increasing number of specific
molecular markers for different diseases, such as oral and breast
cancer, cardiovascular diseases, and human immunodeficiency virus
(HIV), are being identified [12–15]. Hence, salivary proteomics
has been a growing topic of research due to saliva’s ease of collec-
tion and its ability to be a diagnostic tool for systemic diseases.
Saliva can be collected with or without stimulation [16]. Stimulated
and unstimulated saliva produce different flow rates and have dif-
ferent protein compositions and pH [17]. Therefore, it is impor-
tant to use only one type of collection technique when performing
studies. Salivary flow rate can be affected by the degree of hydration
of the subject and olfactory stimulation and also has a circadian
rhythm with peak flow in the afternoon [16]. Therefore, it is
important to control these variables when collecting saliva samples.
Gingival crevicular fluid was first recognized in 1899 as a fluid
that emerges between the surface of the tooth and the epithelial
integument [18]. In healthy sites, GCF production is thought to be
a serum transudate due to the passage of fluid from the surrounding
capillaries into the sulcus [19]. In inflamed or mechanically stimu-
lated sites, GCF production is an inflammatory exudate thought to
act as a protective mechanism to flush away bacteria and transport
antibacterial substances [20]. Gingival crevicular fluid contains
many substances such as blood serum and plasma components,
periodontal epithelium, inflammatory and immune cells, enzymes,
cytokines, interleukins, and products of tissue breakdown [5]. In
healthy sites, the protein concentration is similar to interstitial fluid,
whereas in inflamed sites, GCF protein concentration is similar to
that of serum [21]. Gingival crevicular fluid is an excellent source
for assaying specific dental biomarkers of oral conditions due to its
ability to be site specific (e.g., assaying for abnormalities in dental
tissues in one tooth).
Several techniques have been used to collect GCF [22–24].
The most common and practical method entails the use of absor-
bent filter paper strips to collect the GCF samples. The strip can be
inserted just at the entrance of the crevice, to the base of the pocket,
Biomarkers in Saliva and GCF 551
or until minimum resistance is felt [25]. The advantages of this

technique include being quick and easy to use, being able to apply it
to individual sites, and causing minimal trauma when used cor-
rectly. The amount of GCF collected may be measured by various
means; however, the most common and accurate method used
today is the electronic method. This method measures the amount
of GCF collected on Periopaper® by using an electronic transducer
(Periotron®, Winnipeg, Manitoba). The electronic method is based
on the fact that the flow of electronic current is affected by the
wetness of the paper strip and thus provides a digital readout [26].
Recent advances in proteomic technologies have facilitated the
comprehensive profiling of protein expression in body fluids and
tissues [27]. Applying these proteomic approaches on the readily
accessible saliva samples from patients can aid in early prediction of
the disease status. Indeed, mass spectrometry-based proteomic
approaches have been recently used to identify characteristic saliva
biomarkers for multiple diseases such as Sjögren’s syndrome and
gastric cancer [28, 29].
2 Materials
2.1 Whole Saliva Materials needed:

Sampling and
1. 14 mL polypropylene 17 100 mm Falcon tube.
Preparation
2. Disposable funnel.
3. Wet ice in Styrofoam cup.
4. Distilled water.
5. Aprotinin (1 mg/mL in water).
6. Phenylmethanesulfonyl fluoride (PMSF) (200 mM in MeOH).
7. Labeling machine/label maker for cataloguing tubes (optional;
can use fine-tipped Sharpie marker).
8. Pipettors (for 20 μL and 1000 μL).
9. Pipettor tips (for 20–200 μL and 1000 μL).
10. 1.5 mL microfuge tubes.
2.2 Gingival Materials needed:

Crevicular Fluid
1. Periopaper® gingival fluid collection strips (OraFlow,
Sampling and
Smithtown, NY).
Preparation
2. Clock/timer able to record the seconds.
3. Labtop cooler or dry ice container to place tubes containing
samples.
4. Clinical exam kit with cotton forceps.
5. Cheek retractors and bite block.
6. 2 2 gauze squares.
7. Cotton rolls.
8. Permanent fine-tip marker or label maker.
9. 1.5 mL microfuge tubes.
10. 12 75 mm polystyrene culture tube with caps (Fisher Scien-
tific, Pittsburgh, PA).
11. Pipettors (for 20, 100, 200, and 1000 μL).
12. Pipettor tips (for 20–200 μL and 1000 μL).
13. GCF extraction buffer (24.5 mL phosphate-buffered saline
(pH 7.4), 125 μL phenylmethylsulfonylfluoride (PMSF)
(200 mM in MeOH), 250 μL aprotinin (1 mg/mL in water),
and 83.5 μL of 30% human serum albumin.
2.3 Protein Profiling 1. Dithiothreitol (DTT).

2. Iodoacetamide.
3. Acetone.
4. 2% sodium dodecyl sulfate (SDS).
5. Tris-HCl (pH 8).
6. Qubit fluorometer.
7. 10% Bis-Tris protein mini-gel.
8. MOPS buffer system.
9. Automated in-gel digestion system.
10. Ammonium bicarbonate.
11. Acetonitrile.
12. Trypsin.
13. Formic acid.
14. Nano LC-MS/MS with an HPLC system interfaced to a mass
spectrometer.
3 Methods
3.1 Whole Saliva Carry out all procedures at room temperature unless otherwise
(WS) Collection, specified.
Processing, Storage,
and Preparation
3.1.1 Whole Saliva 1. Prepare tubes for specimen collection:
Collection (a) Label tube with subject information.
(b) Fill Styrofoam cup with wet ice, and place Falcon tube in
cup surrounded by the ice.
(c) Place a disposable funnel into the Falcon tube.
2. Subjects should abstain from brushing their teeth, chewing

gum, eating, or drinking for at least 1.5 h before the visit (see
Note 1).
3. During the collection period, the subject shall be seated
straight up with eyes open and head tilted slightly forward.
4. The subject will hold the Styrofoam cup with the Falcon tube
inside on wet ice during the collection process.
5. The subject will be instructed to minimize orofacial move-
ments to minimize influence on salivary flow (see Note 2).
6. Give the subject distilled drinking water, and ask that they rinse
their mouth out well for 1 min. The subject can then expecto-
rate or swallow the water. Wait 90 s before beginning collection
process. Subject can swallow saliva normally during this time.
7. When starting collection, invite the subject to swallow any
remaining saliva but to abstain from swallowing for the remain-
der of the collection process. The subject is asked to allow the
saliva to accumulate on the floor of the mouth until enough
saliva has pooled, so that they are able to tilt their head forward
to allow the saliva to drip into the funnel.
8. Collection is complete once 3 mL of whole saliva is collected.
This usually takes 10–15 min. Leave the sample on ice, and
perform the processing and storage procedure as soon as pos-
sible (see Note 3). With all the samples still on ice, the protease
inhibitor cocktail containing 80 μM of aprotinin, 4 mM of
bestatin hydrochloride, 1.4 mM of N-trans-epoxysuccinyl-L-
leucine 4-guanidinobutylamide (E-64), 2 mM of leupeptin,
1.5 mM of pepstatin A, and 104 mM of 4-2-aminoethylbenze-
nesulfonyl fluoride hydrochloride (AEBSF) (Sigma-Aldrich,
catalog # P8340, St. Louis, MO) is to be added immediately
chairside to the cell-free WS at a 1:100 dilution to inhibit the
degradation of the proteins in the sample. After vortexing for
30 s, the sample is divided into 500 μL aliquots and stored at
80 C until further analysis.
3.1.2 WS Sample 1. Measure the total volume of saliva within the Falcon tube.
Processing and Storage 2. Per 2 mL of saliva, add the following to inhibit protein degra-
dation (see Note 4):
(a) 20 μL aprotinin (1:100 dilution of 1 mg/mL stock; stored
at 20 C, thaw before use)
(b) 10 μL PMSF (1:200 dilution of 200 mM stock; stored at
20 C but does not freeze)
3. Mix tube well by inverting end over end 10–15 times to mix
completely; then divide into 500 μL aliquots in 6 pre-labeled
1.5 mL microfuge tubes.
4. Transfer tubes into storage boxes and label box accordingly.

5. Make a log sheet with a grid to match each storage box. This
greatly facilitates future sample location and retrieval.
6. Store box in 80 C freezer.
3.1.3 WS Sample 1. When ready for analysis, thaw tube on wet ice.
Preparation 2. Once samples are fully thawed, homogenize samples for 5 s
with an ultrasonic tissue homogenizer (at setting 5). This will
break up the mucous portion of the whole saliva and will make
collection of the supernatant much easier (see Note 5).
3. Centrifuge the samples for 15 min at ~10,000 g at 4 C to
remove insoluble material and obtain a cell-free supernatant.
3.2 Gingival Gingival crevicular fluid (GCF) contains many components that can
Crevicular Fluid (GCF) be used as diagnostic aids. GCF samples are taken from the mesial
Sample Collection, or distal aspect of teeth in humans. The sampling is performed after
Processing, Storage, completing a plaque index score and in order to avoid mechanical
and Preparation irritation and/or bleeding from PD measurements with the peri-
odontal probe. Attention must be given such that the sample is
relatively free of plaque and there is no contamination by saliva and
blood. Carry out all procedures at room temperature unless other-
wise specified.
3.2.1 GCF Collection and 1. Label microfuge tubes accordingly prior to taking samples with
Storage subject info, time point/date, study name/initials, and
sample site.
2. Cheek retractors and a bite block are placed, and the area
around each sample site is isolated using cotton rolls (see
Note 6). The area is dried with gauze and a quick blast of air
from the air/water syringe making sure not to direct any air
flow into the gingival sulcus.
3. Using cotton forceps, hold the orange nylon portion of the
Periopaper® so that the white cellulose portion of the methyl-
cellulose strip can be carefully inserted into the gingival crevice
until a slight resistance is felt (Fig. 1). Each GCF strip will
remain in position for a total of 60 seconds (see Note 7) before
immediate removal (see Notes 8 and 9).
4. Following GCF collection, each strip is placed into its
corresponding labeled microfuge tube (Fig. 2) and is immedi-
ately placed onto dry ice for transport to the laboratory and
storage in a 80 C freezer until analysis (see Notes 10
and 11).
3.2.2 GCF Sample Proteins within the harvested crevicular fluid are extracted from the
Preparation GCF strips using an elution method adapted from Giannobile et al.
[26], involving a series of washes and centrifugations. The elution
Fig. 1 Placement of methylcellulose strips into gingival sulcus. Using cotton

forceps, hold the orange nylon portion of the Periopaper® so that the white
cellulose portion of the methylcellulose strip can be carefully inserted into the
gingival crevice approximately 1 mm or until slight resistance is felt. Optimally
each GCF strip will remain in position for a total of 60 s before removal (The GCF
strip can be removed after 60 s or less (10–30 s), depending of the GCF flow
variables). Because they are site specific, GCF collection studies are very useful
in measuring the levels of dental biomarkers for a particular tooth
Fig. 2 Example of the GCF amount collected in one case. The volume can be
determined using an electronic measuring device, the Periotron® (Winnipeg,
Manitoba), which measures the electrical current flow of the wetted strips
Fig. 3 Securing GCF strip to the side of the culture tube. After adding GCF elution
buffer, the strip is secured at the top of a 12 75 mm polystyrene culture tube
using a cap to hold the orange portion of the strip in place. Ensure that the white
portion of the strip lays flat against the wall of the tube
buffer used for GCF protein extraction is to be made fresh and kept
on wet ice throughout the entire extraction process to inhibit
protease activity (see Note 12).
Procedure:
1. The GCF strips are transported on wet ice from the 80 C
freezer, thawed at room temperature, and kept on wet ice
during the entire procedure.
2. A total of 11 μL of extraction buffer is pipetted directly onto
the cellulose (white) portion of each Periopaper strip. We use
11 μL because there is approximately 1 μL of elution buffer
that will remain on the strip even after centrifugation—leaving
10 μL of washed fluid at the bottom of the tube. The strip is
then secured at the top of a 12 75 mm polystyrene culture
tube using a cap to hold the orange portion of the strip in place.
Ensure that the white portion of the strip is laying flat against
the wall of the tube (Fig. 3).
3. The tubes are then centrifuged at 2000 rpm for 5 min at 4 C.
This process is repeated four additional times to yield 50 μL
total volume.
4. This entire product of 50 μL is then transferred to a sterile
1.5 mL microfuge tube. Each tube is labeled as previously
described, to contain the necessary information.
5. The strips are then soaked in 60 μL of elution buffer and
centrifuged at 2000 rpm for 5 min at 4 C as a final wash to
collect any remaining protein. Approximately 50 μL can be
recovered from this wash and is transferred to the previously

collected 50 μL to yield a total value of 100 μL. The tubes are
then placed in a 80 C freezer until ready for analysis (see
Notes 13–15).
3.3 Protein Profiling Protein profiling of both WS and GCF samples can be conducted
for WS and GCF by custom protein extraction, sodium dodecyl sulfate-
polyacrylamide gel electrophoresis (SDS-PAGE), robotic in-gel
digestion with trypsin, and liquid chromatography-mass spectrom-
etry (LC-MS/MS), followed by database searching and reporting
for protein identification.
3.3.1 Protein Extraction 1. Thaw samples on ice.

2. Remove 500 μL of each sample, and clarify it by centrifugation
at 10,000 g for 10 min.
3. Reduce the sample with 10 mM dithiothreitol at 25 C for
30 min.
4. Alkylate the sample with 15 mM iodoacetamide at 25 C for
5. Concentrate the proteins by adding 4 volumes of 20 C
acetone, and incubate overnight at 20 C.
6. Collect precipitated proteins by centrifuging for 10 min at
5000 g and 4 C.
7. Wash the protein pellets twice with 20 C acetone (see
Note 17).
3.3.2 Protein 1. Suspend the pellets in 55 μL of 2% SDS and 50 mM Tris-HCl.

Quantification, SDS-PAGE, 2. Quantify the protein yield using a Qubit fluorometry assay.
and Automated In-Gel
3. Process 10 μg of each sample by 5 cm SDS-PAGE using a 10%
Digestion
Bis-Tris protein mini-gel and the MOPS buffer system.
4. Excise the mobility region into 10 equally sized bands for
in-gel digestion.
5. Process each band by using a robotic in-gel digestion system
with the following protocol:
(a) Washing with 25 mM ammonium bicarbonate followed
by acetonitrile
(b) Reduction with 10 mM dithiothreitol at 60 C followed
by alkylation with 50 mM iodoacetamide at RT
(c) Digestion with trypsin at 37 C for 4 h
(d) Quenching with formic acid (the supernatant can be ana-
lyzed directly without further processing)
3.3.3 Mass Spectrometry Analyze the tryptic digest by nano LC/MS/MS with an HPLC
and Data Processing system interfaced to a mass spectrometer (we used a nano LC-MS/
MS with a Waters NanoAcquity HPLC system interfaced to a
ThermoFisher Q-Exactive).
Procedure:
1. Load the peptides on a trapping column, and elute them over a
75 μm analytical column at 350 nL/min.
2. Pack both columns with suitable resin (see Note 18).
3. Operate the mass spectrometer in data-dependent mode, with
the Orbitrap operating at 60,000 FWHM and 17,500 FWHM
for MS and MS/MS, respectively.
4. Select the 15 most abundant ions for MS/MS.
5. Using a copy of the resulting Mascot file, search the data with
the following parameters:
Enzyme: Trypsin
Database: Swiss-Prot human (concatenated forward and reverse
plus common potential contaminants)
Fixed modification: Carbamidomethyl (C)
Variable modifications: Oxidation (M), acetyl (N-term), pyro-
Glu (N-term Q), deamidation (N/Q)
Mass values: Monoisotopic
Peptide mass tolerance: 10 ppm
Fragment mass tolerance: 0.02 Da
Max missed cleavages: 2
6. Parse the Mascot DAT files into the scaffold algorithm for
validation, filtering, and creation of a nonredundant list per
sample (see Notes 19 and 20).
4 Notes
1. Due to diurnal variation of analytes in saliva, subjects should

always have their visits scheduled at a given time of day.
2. The subject should not swallow and should not speak during
the collection process.
3. To familiarize the subject with this method, it is wise to run a
1–2 min trial collection prior to the actual collection period.
4. Another protease inhibitor cocktail containing 80 μM of apro-
tinin, 4 mM of bestatin hydrochloride, 1.4 mM of N-trans-
epoxysuccinyl-L-leucine 4-guanidinobutylamide (E-64), 2 mM
of leupeptin, 1.5 mM of pepstatin A, and 104 mM of 4-2-
aminoethylbenzenesulfonyl fluoride hydrochloride (AEBSF)
(Sigma-Aldrich, catalog # P8340, St. Louis, MO) can be used.

20 μL of the cocktail (1:100 dilution of a 1 mg/mL stock;
stored at 20 C, thaw before use) can be added to 2 mL of
saliva. The cocktail must be added immediately chairside to the
cell-free WS at a dilution of 1:100 to inhibit the degradation of
the proteins in the sample. After vortexing for 30 s, the sample
can then be divided into aliquots and stored at 80 C until
further analysis.
5. Be sure to have samples on ice as the homogenization process
will heat the sample up.
6. If present, any supragingival plaque is to be gently removed
prior to sampling.
7. Collection times for GCF samples can vary in the literature.
While the majority of studies reported 30 s, 60 s and 3 min have
been also found in the literature. The reason for this variation
depends on the fluid-flow dynamics of the host. Healthy hosts
have a slower GCF rate, while unhealthy hosts (i.e., periodontal
disease) have faster rates. It is important to note that the
amount of time used for collection should remain constant
throughout a study and constant from patient to patient.
8. Manipulation of the sulcus prior to sampling may create bleed-
ing. It is advised to not manipulate the sulcus prior to taking
GCF samples. If bleeding occurs at the site prior to sampling, it
must be rinsed and cleared away prior to taking the sample.
Blood appearing on the strip can affect the microbiologic
testing results.
9. Multiple samples can be collected at once by sequentially plac-
ing the strips in different sites. This will save time from sam-
pling each site individually and waiting 60 seconds each. For
example, to collect four samples from four different sites (60 s
collection time), a strip can be placed at 0, 15, 30, and 45 s at
site number 1, 2, 3, and 4, respectively. At 60, 75, 90, and
105 s, strips 1, 2, 3, and 4 can be removed, respectively (Fig. 4).
10. Assess the paper strip for over saturation of saliva. If saliva
saturation should occur, wait 90 s and another sample of
GCF may be taken. Retaking of a second sample may occur if
there is a failure to acquire a quality sample on the first attempt.
Examples of this might be that the strip did not fully engage
into the sulcus. If more than one sample in the same site is
necessary, wait 90 s prior to taking another GCF sample. This
allows the GCF flow to return to the sulcus. The same techni-
ques should be utilized on second sampling.
11. It is recommended that since the strips are stored without
stabilizers or protease inhibitors, subsequent analysis is
Fig. 4 Placement of four methylcellulose strips in an orthodontic patient. In this

case, the GCF strips were placed in the sulcus at the mesiobuccal line angles for
a total of 60 s for each maxillary incisor. Four samples here are sampled at once
using the following sequence: at 0-, 15-, 30-, and 45-s intervals, a strip will be
placed at site number 1, 2, 3, and 4, respectively. Another sample was then
retaken at the distobuccal line angles after waiting 90 s between samplings to
allow for GCF flow to return to the sulcus
performed as soon as possible to minimize potential protein

degradation.
12. The elution buffer is best made fresh prior to GCF elution. It is
recommended to make the elution buffer within 24 h of use to
ensure proper elution of the GCF solution. Smaller batches are
made at a time to ensure fresh elution buffer is used. Do not
freeze-thaw the elution buffer for any reason prior to use.
13. GCF elution is best done by eluting one patient at a time, with
all of their time points eluted during the same washing interval.
While eluting one time point in the centrifuge, a second time
point can be eluted by washing at the same time, etc. This
ensures that if for any reason there is a variation in the elution
buffer, the elution buffer composition itself will remain con-
stant for each subject throughout all their time points.
14. Ideally, all patients should be eluted at the same time. How-
ever, due to time constraints, one may wish to do only a few
patients at a time. Minimize the variation in time of elutions for
each patient, and one operator should perform the elution
process to ensure there is no variation in the elution technique.
15. Label each centrifuge tube with tape and a permanent marker
to prevent labels from rubbing off during the elution process,
which may get wet while on ice. Aliquot the final solution
based on the number of protein assay kits you will be running
(i.e., if 200 μL is needed per ELISA kit, aliquot into 200 μL

amounts) to minimize the freeze-thaw cycles of each solution.
16. It is recommended to incubate the samples in the dark after
alkylation.
17. We also recommend air-drying the pellets to remove residual
acetone.
18. We recommend packing with Jupiter Proteo resin
(Phenomenex).
19. Filter the data using 1% protein and peptide FDR and with at
least two unique peptides required per protein.
20. An example of detailed protein identification data can be found
in an Excel workbook accompanying this chapter, where we
identified the proteins from saliva samples acquired from 10 cli-
ents (numbered: 22640-22649) and divided into 2 groups: RA
group (224640-44) and CO group (22645-49). The Excel file
contains four worksheets:
– Two protein report worksheets that contain the full list of
proteins identified (including known contaminants and
reverse hits) and their molecular weight and spectral counts
(SpC) with a summary of the data
– An NSAF Calc. sheet that contains the conversion to spec-
tral abundance factor (SAF) and subsequent normalized
spectral abundance factor (NSAF). This was based on the
following equation:
NSAF ¼ ðSpC=MW Þ=ΣðSpC=MW ÞN

where SpC is spectral counts, MW is protein MW in kDa,
and N is total number of proteins.
– T-test sheet that contains SpC and NSAF values, the average
of the NSAF values for the two groups RA and CO, and also
a T-test performed using the NSAF data.
References
1. Loo JA, Yan W, Ramachandran P, Wong DT 4. Levine M (2013) Salivary proteins may be use-
(2010) Comparative human salivary and ful for determining caries susceptibility. J Evid
plasma proteomes. J Dent Res 89:1016–1023 Based Dent Pract 13:91–93
2. Kinney JS, Ramseier CA, Giannobile WV 5. Balducci L, Ramachandran A, Hao J et al
(2007) Oral fluid-based biomarkers of alveolar (2007) Biological markers for evaluation of
bone loss in periodontitis. Ann N Y Acad Sci root resorption. Arch Oral Biol 52:203–208.
1098:230–251 https://doi.org/10.1016/j.archoralbio.2006.
3. Lamster IB, Ahlo JK (2007) Analysis of gingi- 08.018
val crevicular fluid as applied to the diagnosis of 6. Kereshanan S, Stephenson P, Waddington R
oral and systemic diseases. Ann N Y Acad Sci (2008) Identification of dentine sialoprotein
1098:216–229 in gingival crevicular fluid during physiological
root resorption and orthodontic tooth move- Anat Rec 162:313–325. https://doi.org/10.
ment. Eur J Orthod 30:307–314. https://doi. 1002/ar.1091620306
org/10.1093/ejo/cjn024 19. Pashley DH (1976) A mechanistic analysis of
7. Mandel ID, Wotman S (1976) The salivary gingival fluid production. J Periodontal Res
secretions in health and disease. Oral Sci Rev 11:121–134. https://doi.org/10.1111/j.
8:25–47 1600-0765.1976.tb00060.x
8. Kaufman E, Lamster IB (2000) Analysis of 20. Brill N (1959) Removal of particles and bacte-
saliva for periodontal diagnosis. A review. J ria from gingival pockets by tissue fluid. Acta
Clin Periodontol 27:453–465. https://doi. Odontol Scand 17:431–440. https://doi.org/
org/10.1034/j.1600-051x.2000. 10.3109/00016355908993933
027007453.x 21. Griffiths GS (2003) Formation, collection and
9. Gemmell E, Marshall RI, Seymour GJ (1997) significance of gingival crevice fluid. Periodon-
Cytokines and prostaglandins in immune tol 2000 31:32–42
homeostasis and tissue destruction in peri- 22. Skapski H, Lehner T (1976) A crevicular wash-
odontal disease. Periodontol 2000 ing method for investigating immune compo-
14:112–143. https://doi.org/10.1111/j. nents of crevicular fluid in man. J Periodontal
1600-0757.1997.tb00194.x Res 11:19–24. https://doi.org/10.1111/j.
10. Kaufman E, Lamster IB (2002) The diagnostic 1600-0765.1976.tb00046.x
applications of saliva—a review. Crit Rev Oral 23. Sueda T, Bang J, Cimasoni G (1969) Collec-
Biol Med 13:197–212 tion of gingival fluid for quantitative analysis. J
11. Rathnayake N, Akerman S, Klinge B et al Dent Res 48:159. https://doi.org/10.1177/
(2013) Salivary biomarkers for detection of 00220345690480011501
systemic diseases. PLoS One 8:e61356. 24. Brill N, Krasse BO (1958) The passage of tissue
https://doi.org/10.1371/journal.pone. fluid into the clinically healthy gingival pocket.
0061356 Acta Odontol Scand 16:233–245. https://doi.
12. Miller CS, King CP, Langub MC et al (2006) org/10.3109/00016355809064110
Salivary biomarkers of existing periodontal dis- 25. Lamster IB, Oshrain RL, Fiorello LA et al
ease: a cross-sectional study. J Am Dent Assoc (1988) A comparison of 4 methods of data
137:322–329. https://doi.org/10.14219/ presentation for lysosomal enzyme activity in
JADA.ARCHIVE.2006.0181 gingival crevicular fluid. J Clin Periodontol
13. Boyle JO, Mao L, Brennan JA et al (1994) 15:347–352. https://doi.org/10.1111/j.
Gene mutations in saliva as molecular markers 1600-051X.1988.tb01010.x
for head and neck squamous cell carcinomas. 26. Giannobile WV, Lynch SE, Denmark RG et al
Am J Surg 168:429–432 (1995) Crevicular fluid osteocalcin and pyridi-
14. Hu S, Arellano M, Boontheung P et al (2008) noline cross-linked carboxyterminal telopep-
Salivary proteomics for oral cancer biomarker tide of type I collagen (ICTP) as markers of
discovery. Clin Cancer Res 14:6246–6252. rapid bone turnover in periodontitis: a pilot
https://doi.org/10.1158/1078-0432.CCR- study in beagle dogs. J Clin Periodontol
07-5037\r14/19/6246[pii] 22:903–910. https://doi.org/10.1111/j.
15. Zhang L, Xiao H, Karlan S et al (2010) Discov- 1600-051X.1995.tb01793.x
ery and preclinical validation of salivary tran- 27. Hu S, Jiang J, Wong DT (2010) Proteomic
scriptomic and proteomic biomarkers for the analysis of saliva: 2D gel electrophoresis,
non-invasive detection of breast cancer. PLoS LC-MS/MS, and Western Blotting. Methods
One 5:e15573. https://doi.org/10.1371/ Mol Biol (Clifton, NJ) 666:31–41. https://
journal.pone.0015573 doi.org/10.1007/978-1-60761-820-1_3
16. NAVAZESH M (1993) Methods for collecting 28. Peluso G, De Santis M, Inzitari R et al (2007)
saliva. Ann N Y Acad Sci 694:72–77. https:// Proteomic study of salivary peptides and pro-
doi.org/10.1111/j.1749-6632.1993. teins in patients with Sjögren’s syndrome
tb18343.x before and after pilocarpine treatment. Arthri-
17. Golatowski C, Gesell Salazar M, Dhople VM tis Rheum 56:2216–2222. https://doi.org/
et al (2013) Comparative evaluation of saliva 10.1002/art.22738
collection methods for proteome analysis. Clin 29. Xiao H, Zhang Y, Kim Y et al (2016) Differen-
Chim Acta 419:42–46. https://doi.org/10. tial proteomic analysis of human saliva using
1016/j.cca.2013.01.013 tandem mass tags quantification for gastric can-
18. Bevelander G, Nakahara H (1968) The fine cer detection. Sci Rep 6. https://doi.org/10.
structure of the human peridental ligament. 1038/srep22165
INDEX
A D
Ameloblastin (AMBN) .............................. 169, 219–227, Demineralization-remineralization
229–235, 252 protocol.............................................................. 380
Ameloblastin mRNAs.................................................... 163 Dental applications............................................... 121–127
Ameloblasts ..................................................... v, 4, 10, 33, Dental epithelial stem cell (DESCs) ...........................3, 4,
169, 182, 220, 226, 230, 251, 293 8–10, 92
Amelogenesis .............................................. 130, 230, 251, Dental mesenchymal stem cells (DMSCs)...............59–74
253, 267, 409, 410, 450, 454, 458 Dental mineralized tissues ...........................173–180, 357
Amelogenin ................................................ 129, 137, 144, Dental papilla ............................................................13–18
182, 219–225, 229, 248, 252, 253, 260–262, 264 Dental papilla mesenchymal cell line .......................13–18
Antisense RNA probes................................ 163, 182, 184 Dental pulp (DP) ....................23, 30, 33–35, 44, 60, 77,
Apatite..................................................129, 135, 136, 370 80, 82, 94, 142, 328, 394, 398, 400, 402, 489
Apical papilla................................................ 60, 62, 63, 80 Dental pulp mesenchymal stem cells
Apical papilla stem cells.............................................59–73 (DP-MSCs)....................................................77–89
Artifact reduction .......................................................... 311 Dental pulp stem cells (DPSCs)....................... 13, 21–26,
Artificial dental implants ............................................... 139 59–74, 91–97, 109
Artificial saliva............................................. 130, 132, 133, Dental stem cells (DSCs)..................................... v, 29–36,
136, 383, 390, 391 59, 60, 62–64, 66, 70, 72
Autofluorescence.................................179, 191, 192, 224 Dental tissue repair ....................................................... 121
Dental tissues.......................................................... v, 6, 10,
B 29, 50, 92, 136, 139, 163, 175, 176, 178, 182,
Background interference .............................................. 103 189, 198, 239–249, 453–490, 550
Dentin............................................................ v, 13, 14, 22,
Bioengineered dental tissues ............................... 139, 140
Bioengineered Tooth Bud (BTB) model............ 139–149 29, 30, 39, 221, 271, 307, 314, 357, 372, 388,
Bone morphogenetic protein 2 (Bmp2)..................13–18 395, 410, 454, 549
Bone morphogenetic protein 2 knock out Dentin extracellular matrix (ECM)................................ 92
Dentin-forming odontoblasts ........................................ 21
mouse (Bmp2 cKO) ............................................ 13
Bone sialoprotein (BSP) ............................................... 217 Dentin matrix protein 1 (DMP1) ................................ 217
Dentinogenesis .....................................................v, 13, 30,
C 92, 111–118, 409
Dentinogenesis imperfecta (DGI) ..............................410,
Cell cultures.............................................. 3, 4, 14, 23, 26, 455, 457, 458
60, 70, 94, 99, 105–108, 126, 153, 192, 456, 478 Dentin sialophosphoprotein (DSPP)..........................144,
Cell lineage tracing ............................................... v, 39–45 240, 249, 315, 458
Cell line establishment and characterization ....................v Dentin tubules ....................................14, 39, 43–45, 117
Cell progeny ..............................................................40, 44 Developing mouse molars ............................................ 219
Confocal laser scanning microscopy (CLSM) ............103, Differentiations ........................................................ v, 3, 4,
104, 106–108, 363, 370, 374, 375 13, 29, 30, 39, 40, 44, 59, 61, 67–72, 78, 92, 93,
Continuously growing mouse incisors .....................3–10, 96, 99, 111, 139, 293, 395
29–36, 314 Digital 3D datasets of whole embryos......................... 198
Contrast agent...................................................... 312, 316 DIG-labeled RNA probes................... 151–158, 163–171
Craniofacial development .................................... 163, 454 Digoxigenin (DIG) labeling........................................152,
Cre fluorescent protein (GFP) ...........14, 16, 24, 25, 108 154, 164, 166, 184
Cre recombinase................................................. 14–16, 39 Drug delivery system .................................................... 121
Cryostat sections ........................................................... 176
https://doi.org/10.1007/978-1-4939-9012-2, © Springer Science+Business Media, LLC, part of Springer Nature 2019
563
ODONTOGENESIS: METHODS AND PROTOCOLS
564 Index
E In vitro and in vivo models ..................................... vi, 111
In vitro tooth mineralization............................... 129, 137
Enamel In vivo quantitative co-localization..................... 219, 220
biomimetics .................................................... 133–136 In vivo tooth bud model implantation ........................ 144
matrix proteins ................. v, 219, 220, 230, 251–264 Ion-exchange chromatography .................. 249, 252, 253
regrowth ......................................................... 129–137 Ion-exchange fast protein liquid chromatography
remineralization strategies ...................................... 129 (FPLC)............................................................... 211
Epithelial-mesenchymal interactions................... 3, 49–53 Isolation of dental stem cell enriched
Epithelial stem cells (ESCs).........................29, 30, 33–34 population......................................................29–36
Establishment of stable cell lines .......................... v, 21–26
Extracellular matrix (ECM)....................................... v, 35, K
68, 92, 219, 229, 230, 325, 326, 357, 358
Kidney capsule transplantation.................................49–53
F
L
FACS sorting.............................................................30, 36
FGF8.............................................................................. 451 Lentivirus ...............................................14, 16, 21, 24, 25
Flow cytometry ................. 36, 66, 67, 77–89, 93, 96, 98 Leucine-rich amelogenin peptide
Fluorescent antibody cell labeling.................................. 16 (LRAP)..............................................130–135, 137
Ligand...........................................................103–110, 252
G Lineage commitment ................................................13, 35
LNA/DNA probes .............................182, 184, 187, 189
Gelatin methacrylate (GelMA) Hydrogels ......v, 139–149 Low affinity nerve growth factor receptor
Gene expression pattern visualization in mouse p75NTR/CD271................................................ 77
embryo ............................................................... 151
Genetic recombination ................................................... 21 M
Genetic transduction....................................................... 21
Gene transfer techniques ................................................ 21 Manders’ coefficient.................................... 220, 225, 226
Gli1-CreERT2 ............................................... 41, 43, 45, 46 Melanoma cell adhesion molecule
Guanidine-HCl/EDTA .............................. 212, 213, 216 MCAM/CD146.................................................. 77
Mesenchymal stem cell antigen MSCA-1...................... 77
H Mesenchymal stem/stromal cells (MSCs) .................... 29,
30, 34, 35, 59–74, 77–89, 93, 96
Human embryonic kidney cells 293 Micro-computed tomography (μCT) .........................116,
(HEK 293 cells) .................................................. 21 198, 309–320
Human tooth enamel ................................................... 135 Microdissection ......................................... 3–10, 334, 396
Hydrogel scaffolds ........................................................ 139 Microscopy techniques .......................103–105, 132, 363
Mineralization-induced conditions .............................. 103
I
Molar damage....................................................... 113, 114
Immortalized deleted Bmp2 dental papilla mesenchymal Mouse embryo fixation and staining .................. 203, 205
(iBmp2ko/ko-dp) cell line ...........................16–18 Mouse embryos................... 51, 151–158, 163, 199, 203
Immortalized mouse floxed Bmp2 dental papilla Mouse incisors.............................. 3–10, 29–36, 312, 314
mesenchymal (iBmp2flox/flox-dp) Mouse model........................................................ 111–118
cells.................................................................13–18 mRNAs ................................................182, 184, 252, 539
Immunocompromised mice ................................ 140, 144 Multiparametric flow cytometry...............................77–89
Immunocytochemistry (ICC) ............................ 106, 108, Multiwalled carbon nanotubes (MWNTs) ......v, 121–127
110, 325, 329 Murine developing enamel tissue........................ 193–195
Immunofluorescence (IF)............................. v, 39–45, 98,
142, 148, 191–195 N
Immunohistochemistry (IHC)....................... v, 144, 151, Next-generation dental material .................................. 129
173, 179, 402 Non-collagenous proteins (NCPs) .............................211,
Immuno-labeling .......................................................... 221 212, 215–218, 240
In situ hybridization (ISH) .............................v, 151–158, Non-specific staining and false positive ....................... 191
163–171, 181–190 Nucleotide locked-nucleic acid (LNA)-incorporated
In situ protein Visualization ................................ 173–180 oligodeoxynucleotide probes (20 LNA/
Intramolecular signaling visualization ................ 103–110 DNA nucleotide probes) .................................. 181
ODONTOGENESIS: METHODS AND PROTOCOLS
Index 565
O S
Odontoblast .................................................. v, 21, 22, 40, Scaffolds......................................91, 93, 96, 99, 139, 558
43–47, 92, 176, 395 Self-renewal properties ................................................... 77
differentiation.....................................................14, 30, Silver staining ......................................201, 209, 260, 261
39, 91–97 Small Integrin-Binding Ligand, N-linked
lineage ........................................................... 13, 30, 35 Glycoprotein (SIBLING) family ...................... 455
specific markers.......................................................... 93 Somatic stem cells ........................................................... 21
Odontogenesis ...................................................... v, vi, 93, Sonic hedgehog mRNA................................................... 163
309, 453, 454 Sox2-GFP+ cell population ............................................ 30
Odontogenic progenitors ............................................... 92 Specificity of in situ RNA detection.................... 181–190
Organogenesis ................................................................. 14 Stains-all staining........................ 211, 213, 215–217, 246
Osteopontin (OPN).......................................93, 211, 217 Stem cell driven regenerative medicine.......................... 59
Stem cell-enriched population .................................29–36
P Stem cells ........................................................... 3, 4, 8–10,
13, 21–26, 29–36, 59–74, 78, 91–97, 103, 109
Paraffin-embedded sections.......................................... 144
Peptide-mediated biomimetic enamel regrowth Stro-1 antigen ................................................................. 67
in situ ........................................................ 129–135 Sudan Black B (SBB) ...................................191, 193–195
Synchrotron micro-computed tomography ...............202,
Peptide-mediated remineralization of artificial
dental lesions ..................................................... 130 207, 208
Periodontal ligament .............................. 59–73, 103, 105
T
Periodontal ligament stem cells (PDLSCs) ............59–74,
103–110 3D image analysis software.................................. 198, 202
Phosphorylated glycoprotein ....................................... 211 3D imaging.................................................................... 341
Plasma membrane visualization.................................... 104 3D printing.................................................. 92, 93, 97, 99
Polyethylene glycol (PEG) .................................. 122–126 Tissue engineering .......................................59–74, 91–97
Post-hybridization signal detection ............................. 151 Tissue recombination............................................ v, 49–53
Primary cell line.........................................................21–26 Tissue regeneration ............................. v, 59, 92, 103, 105
Primary dental cell line ................................................... 14 Tissue-Tek O.C.T. compounds ........................... 174, 178
Primary human dental pulp stem cells.....................21–26 Tooth demineralization ...................................... 361, 380,
Primary teeth.............................................................59–73 387, 389, 394, 526
Primary teeth stem cells............................................59–74 Tooth development................................................ v, 3, 13,
Principle of incident angles .......................................... 103 14, 78, 139, 326, 453–455
Proteinase K treatment ............................... 174, 179, 186 Tooth epithelial lineages ................................................. 30
Protein delivery system ................................................. 121 Tooth organ visualization............................................. 197
Protein fractions .................................................. 211, 216, Tooth regeneration ..................................................... v, 29
217, 253, 255, 256, 261, 262, 264 Tooth repair............................................................ 77, 111
Protein immobilization........................................ 124, 125 Tooth tissue engineering .............................................. 139
Total internal reflection microscopy (TIRFM) ..........104,
R 106, 108–109
2D and 3D cultures ........................................................ 93
R26RTomato ........................................................ 41, 43, 45
Refractive index ............................................................. 104
U
Regenerative medicine .................................................... 59
Regenerative therapies for injured dentin ..................... 22 Ultrafiltration ...................................................... 188, 211,
Reparative dentinogenesis mouse model............ 111–118 212, 214, 216, 218, 264
Reporter activation ...................................................39, 46
Restorative dentistry ..................................................... 369 V
RNA detection ..................................................... 181–190 Virtual histology............................................................ 197
RNA probes......................................................... 151–158,
163–171, 182, 184, 186, 187, 189 W
RNAse-free .......................................................... 152–155,
164–166, 168, 170, 186, 188, 532, 538 Whole-mount in situ hybridization
(WMISH) ................................................. 151–158

ODONTOGENESIS

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ODONTOGENESIS

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 1922

Petros Papagerakis Editor

For further volumes:

Methods and Protocols

ISSN 1064-3745 ISSN 1940-6029 (electronic)

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Ann Arbor, MI, USA Petros Papagerakis

PART I PROTOCOLS FOR ESTABLISHMENT OF DENTAL CELLS LINES

PART II PROTOCOLS FOR DENTAL STEM CELLS AND TISSUE REGENERATION

7 Dental Mesenchymal Stem Cells: Dental Pulp Stem Cells, Periodontal

10 In Vitro Analysis of Intramolecular Signaling Events in PDLSCs

PART III PROTOCOLS FOR STUDYING GENE AND PROTEIN EXPRESSION

16 In Situ Hybridization on Mouse Paraffin Sections Using DIG-Labeled

PART IV PROTOCOLS FOR BIOCHEMISTRY AND IMAGING

25 Purification of Developing Enamel Matrix Proteins Using Preparative

PART V PROTOCOLS TO STUDY DENTAL DISEASES

30 Rodent Dental Fluorosis Model: Extraction of Enamel Organ

PART VI PROTOCOLS FOR GENETIC, EPIGENETIC AND CLINICAL STUDIES

36 Protocol GenoDENT: Implementation of a New NGS Panel

37 Protocols for Genetic and Epigenetic Studies of Rare Diseases

ANA CAROLINA ACEVEDO  Oral Histopathology Laboratory, Department of Dentistry, Health

AGNÈS BLOCH-ZUPAN  Faculté de Chirurgie Dentaire, Université de Strasbourg, Strasbourg,

CAROLINE LOURENÇO DE LIMA  Oral Histopathology Laboratory, Department of Dentistry,

CARLOS GONZÁLEZ-CABEZAS  Department of Cariology, Restorative Sciences and

CHAOYUAN LI  Department of Biomedical Sciences, Texas A&M University College of

KEISHI OTSU  Division of Developmental Biology and Regenerative Medicine, Department

MARI M. SAITO  Department of Biochemistry and Molecular Biology, School of Dental

KOSTAS VERDELIS  Department of Restorative Dentistry/Comprehensive Care, School of

Protocols for Establishment of Dental Cells Lines and Animal

Microdissection and Isolation of Mouse Dental Epithelial

Tooth development proceeds with sequential and reciprocal epithe-

Rodent incisors are known to be continuously growing teeth

2.1 Incisor All materials should be sterile.

Fig. 1 Stereomicroscope for the separation of incisors from mandible. Overall

2.2 Media 1. Working medium: DMEM/F12 supplemented with 50 U/mL

2.3 General 1. Humidified incubator at 37  C and 5% CO2.

All procedures must be performed under sterile techniques with

3.1 Extraction of an 1. Rinse surgical equipment in 70% ethanol prior to use.

dental papilla cells

Labial cervical loop epithelium

4. Use forceps and scissors to remove the mouse heads.

Fig. 4 Appearance of an incisor germ after collagenase treatment. Arrow

Fig. 5 Process of separation of a labial epithelial cell sheet

Fig. 6 Appearance of a labial cervical loop. An arrow indicates a labial cervical

3. After the passage several times, the cells become to proliferate

Fig. 9 Appearance of sphere of immortalized dental epithelial stem cells on

1. Mouse incisors of postnatal days 3–7 are the most appropriate

This work was supported, in part, by KAKENHI (26670805 to

Establishment of an Immortalized Mouse Bmp2 Knockout

Tooth development involves sequential and reciprocal interactions

odontoblast differentiation, abnormal dentin tubules, and

2.2 Animals C57BL/6 mice.

2.3 Cells and Cell Embryonic stem (ES) cells of mouse.

2.4 Virus for Lentivirus carrying the SV 40 large T antigen gene.

2.6 Antibodies Primary anti-SV40 large T antigen monoclonal antibody.

2.7 Other Materials Stereomicroscope.

2. GFP-positive cells were observed under a Nikon inverted fluo-

1. The dental papilla mesenchymal cells were grown with alpha

for 4 h. The cells were treated with a mouse monoclonal anti-

We are grateful to the core facility center at the University of Texas

1. Linde A, Goldberg M (1993) Dentinogenesis. 3. Urist MR (1965) Bone: formation by autoin-

Establishment of Stable Cell Lines from Primary Human

Lentiviral expression vector systems have been used extensively to

ANA CAROLINA ACEVEDO Oral Histopathology Laboratory, Department of Dentistry, Health

AGNÈS BLOCH-ZUPAN Faculté de Chirurgie Dentaire, Université de Strasbourg, Strasbourg,

CAROLINE LOURENÇO DE LIMA Oral Histopathology Laboratory, Department of Dentistry,

CARLOS GONZÁLEZ-CABEZAS Department of Cariology, Restorative Sciences and

CHAOYUAN LI Department of Biomedical Sciences, Texas A&M University College of

KEISHI OTSU Division of Developmental Biology and Regenerative Medicine, Department

MARI M. SAITO Department of Biochemistry and Molecular Biology, School of Dental

KOSTAS VERDELIS Department of Restorative Dentistry/Comprehensive Care, School of

2.3 General 1. Humidified incubator at 37 C and 5% CO2.

3.2 Sample 1. Incubate the slides for 10 min in a 37 C chamber to remove